Holger Lange

Head of Department/Head of Research

(+47) 900 88 460
holger.lange@nibio.no

Place
Ås H8

Visiting address
Høgskoleveien 8, 1433 Ås

To document

Abstract

Clumping index (CI) is a measure of foliage aggregation relative to a random distribution of leaves in space. The CI can help with estimating fractions of sunlit and shaded leaves for a given leaf area index (LAI) value. Both the CI and LAI can be obtained from global Earth Observation data from sensors such as the Moderate Resolution Imaging Spectrometer (MODIS). Here, the synergy between a MODIS-based CI and a MODIS LAI product is examined using the theory of spectral invariants, also referred to as photon recollision probability (‘p-theory’), along with raw LAI-2000/2200 Plant Canopy Analyzer data from 75 sites distributed across a range of plant functional types. The p-theory describes the probability (p-value) that a photon, having intercepted an element in the canopy, will recollide with another canopy element rather than escape the canopy. We show that empirically-based CI maps can be integrated with the MODIS LAI product. Our results indicate that it is feasible to derive approximate p-values for any location solely from Earth Observation data. This approximation is relevant for future applications of the photon recollision probability concept for global and local monitoring of vegetation using Earth Observation data.

Abstract

Horizontal Visibility Graphs (HVGs) are a recently developed method to construct networks based on time series. Values (the nodes of the network) of the time series are linked to each other if there is no value higher between them. The network properties reflect the nonlinear dynamics of the time series. For some classes of stochastic processes and for periodic time series, analytic results can be obtained for the degree distribution, the local clustering coefficient distribution, the mean path length, and others. HVGs have the potential to discern between deterministic-chaotic and correlated-stochastic time series. We investigate a set of around 150 river runoff time series at daily resolution from Brazil with an average length of 65 years. Most of the rivers are exploited for power generation and thus represent heavily managed basins. We investigate both long-term trends and human influence (e.g. the effect of dam construction) in the runoff regimes (disregarding direct upstream operations). HVGs are used to determine the degree and distance distributions. Statistical and information-theoretic properties of these distributions are calculated: robust estimators of skewness and kurtosis, the maximum degree occurring in the time series, the Shannon entropy, permutation complexity and Fisher Information. For the latter, we also compare the information measures obtained from the degree distributions to those using the original time series directly, to investigate the impact of graph construction on the dynamical properties as reflected in these measures. We also show that a specific pretreatment of the time series conventional in hydrology, the elimination of seasonality by a separate z-transformation for each calendar day, changes long-term correlations and the overall dynamics substantially and towards more random behaviour. Moreover, hydrological time series are typically limited in length and may contain ties, and we present empirical consequences and extensive simulations to investigate these issues from a HVG methodological perspective. Focus is on one hand on universal properties of the HVG, common to all runoff series, and on site-specific aspects on the other.

To document

Abstract

National parks are established to reduce human influence on nature and contribute to species conservation, biodiversity and ecological services. Other states of protection like the UNESCO world heritage sites, for example, are created for maintaining culturally important places or lifestyles. In the Matobo Hills (Zimbabwe) both states of protection are present, a national park and a world heritage site. In addition, the land outside the National Park belongs to two different systems of ownership, namely “common” (i.e. community-owned) and “not-common” (privately or governmentally owned) land. In this paper, we investigated how the state of protection and the ownership affected the land use and land cover. We derived maps using Landsat images from 1989, 1998 and 2014 by supervised classification with Random Forests. To compensate for the lack of ground data we inferred past land use and land cover from recent observations combining photographs, Google Earth images and change detection. We could identify four classes, namely shrub land, forest, patchy vegetation and agricultural area. The Matobo National Park showed a stable composition of land cover during the study period and the main changes were observable in the surroundings. Outside the national park, forest increased by about 7%. The common lands have changed substantially and their agricultural area decreased. We attribute this development to the Fast Track Land Reform, which took place in the early 2000s. Our approach shows that combining information on recent land cover with change detection allows to study the temporal development of protected areas.

To document

Abstract

Background and aims Decomposition of the finest harvest residues is important for the carbon and nutrient cycle in forest ecosystems both before and after tree felling. We assumed that decomposition is dependent on harvest residue fraction and chemistry, soil temperature and moisture, and aimed at determining decomposition rates and nutrient dynamics of needles, twigs and fine roots from newly felled Picea abies trees in two sites with different climate and topography. Methods Decomposition of needles, twigs and fine roots in mesh bags was followed for up to six years and four years in harvesting sites in eastern and western Norway, respectively. The western site had a more humid climate and a steeper terrain than the eastern site. Results The mass loss after two years was significantly higher for needles (49–59%) than for twigs and fine roots (29–38%). Between sites, there was no significant difference between mass loss for neither needles nor twigs. Nitrogen accumulated in needles during the first year, but 35% of initial needle N had been released after three years. The initial needle and twig decomposition rate was dependent on soil moisture and topographic aspect. During the first three years, needle lignin concentrations retarded whereas P concentrations stimulated needle mass loss. For twigs, P concentrations stimulated mass loss, whereas higher soil temperatures reduced it. Conclusions Lignin and P concentrations of plant parts and soil temperature were the most important factors for the first three-year mass loss. The slow release of nutrients shows the importance of remaining needles, twigs and fine roots as a long-time nutrient source in the ecosystems under study.

Abstract

Short-term (three to four years) effects of forest harvesting on soil solution chemistry were investigated at two Norway spruce sites in southern Norway, differing in precipitation amount and topography. Experimental plots were either harvested conventionally (stem-only harvesting, SOH) or whole trees, including crowns, twigs and branches were removed (whole-tree harvesting, WTH), leaving residue piles on the ground for some months before removal. The WTH treatment had two sub-treatments: WTH-pile where there had been piles and WTH-removal, from where residues had been removed to make piles. Increased soil solution concentrations of NO3–N, total N, Ca, Mg and K at 30 cm depth, shown by peaks in concentrations in the years after harvesting, were found at the drier, less steep site in eastern Norway after SOH and WTH-pile, but less so after WTH-removal. At the wetter, steeper site in western Norway, peaks were often observed also at WTH-removal plots, which might reflect within-site differences in water pathways due largely to site topography.

Abstract

For å verifisere beregningsmetoder og modeller for endringer av karbonlagre i skogsjord brukt i UNFCCC rapporteringen er det behov for data fra jordprøver med gjentak over tid og der prøvetakingsmetoder er konsistente. I Norge finnes ikke denne typen data på nasjonalt / landsdekkende nivå. For å møte behovet for verifisering av beregningsmetodikken brukt i UNFCCC rapporteringen innenfor rammene av eksisterende data med metodisk konsistens over tid, er det i denne undersøkelsen gjennomført ny prøvetaking av jord og vegetasjon i to etablerte forsøksfelt i skog i sør-øst Norge der vi fra før av både har data fra tidligere jordprøveanalyser og tilvekstdata for trær i skogsbestand. De to forsøksfeltene ligger på Nordmoen i Akershus (etablert 1973 og tilplantet 1974) og i Skiptvet i Østfold (etablert 1976 i eksisterende foryngelse med supplerplanting i 1977). Med nye jordprøver, biomassemålinger og vegetasjonsanalyser i 2011 gir dette to tidsserier på hhv. 38 og 34 år med hensyn på endringer i jordkarbon og inngangsverdier i beregningsmodellene. Den eksperimentelle behandlingen i Skiptvet omfatter ulik grad av treslagsblanding av bjørk og gran på de enkelte forsøksrutene, mens på Nordmoen sammenliknes rene bestand av hhv. bjørk, gran og furu. De klimatiske forhold er tilnærmet like, mens jordsmonntypen er ulik med næringsfattig sandjord på Nordmoen og næringsrik leirjord i Skiptvet. Resultatene fra forsøkene er begrenset til å representere klimatiske og vegetasjonsmessige forhold på Østlandet (og forhold tilsvarende de to lokalitetene), og forsøksfeltene er dermed ikke representative eksempelvis for kystnære og kontinentale strøk.

To document

Abstract

Daily precipitation extremes and annual totals have increased in large parts of the global land area over the past decades. These observations are consistent with theoretical considerations of a warming climate. However, until recently these trends have not been shown to consistently affect dry regions over land. A recent study, published by Donat et al. (2016), now identified significant increases in annual-maximum daily extreme precipitation (Rx1d) and annual precipitation totals (PRCPTOT) in dry regions. Here, we revisit the applied methods and explore the sensitivity of changes in precipitation extremes and annual totals to alternative choices of defining a dry region (i.e. in terms of aridity as opposed to precipitation characteristics alone). We find that (a) statistical artifacts introduced by data pre-processing based on a time-invariant reference period lead to an overestimation of the reported trends by up to 40 %, and that (b) the reported trends of globally aggregated extremes and annual totals are highly sensitive to the definition of a "dry region of the globe". For example, using the same observational dataset, accounting for the statistical artifacts, and based on different aridity-based dryness definitions, we find a reduction in the positive trend of Rx1d from the originally reported +1.6 % decade−1 to +0.2 to +0.9 % decade−1 (period changes for 1981–2010 averages relative to 1951–1980 are reduced to −1.32 to +0.97 % as opposed to +4.85 % in the original study). If we include additional but less homogenized data to cover larger regions, the global trend increases slightly (Rx1d: +0.4 to +1.1 % decade−1), and in this case we can indeed confirm (partly) significant increases in Rx1d. However, these globally aggregated estimates remain uncertain as considerable gaps in long-term observations in the Earth's arid and semi-arid regions remain. In summary, adequate data pre-processing and accounting for uncertainties regarding the definition of dryness are crucial to the quantification of spatially aggregated trends in precipitation extremes in the world's dry regions. In view of the high relevance of the question to many potentially affected stakeholders, we call for a well-reflected choice of specific data processing methods and the inclusion of alternative dryness definitions to guarantee that communicated results related to climate change be robust.

Abstract

We investigate a set of long-term river runoff time series at daily resolution from Brazil, monitored by the Agencia Nacional de Aguas. A total of 150 time series was obtained, with an average length of 65 years. Both long-term trends and human influence (water management, e.g. for power production) on the dynamical behaviour are analyzed. We use Horizontal Visibility Graphs (HVGs) to determine the individual temporal networks for the time series, and extract their degree and their distance (shortest path length) distributions. Statistical and information-theoretic properties of these distributions are calculated: robust estimators of skewness and kurtosis, the maximum degree occurring in the time series, the Shannon entropy, permutation complexity and Fisher Information. For the latter, we also compare the information measures obtained from the degree distributions to those using the original time series directly, to investigate the impact of graph construction on the dynamical properties as reflected in these measures. Focus is on one hand on universal properties of the HVG, common to all runoff series, and on site-specific aspects on the other. Results demonstrate that the assumption of power law behaviour for the degree distribtion does not generally hold, and that management has a significant impact on this distribution. We also show that a specific pretreatment of the time series conventional in hydrology, the elimination of seasonality by a separate z-transformation for each calendar day, is highly detrimental to the nonlinear behaviour. It changes long-term correlations and the overall dynamics towards more random behaviour. Analysis based on the transformed data easily leads to spurious results, and bear a high risk of misinterpretation.

To document

Abstract

Global warming will most likely lead to increased drought stress in forest trees. We wanted to describe the adaptive responses of fine roots and fungal hyphae, at different soil depths, in a Norway spruce stand to long-term drought stress induced by precipitation exclusion over two growing seasons. We used soil cores, minirhizotrons and nylon meshes to estimate growth, biomass and distribution of fine roots and fungal hyphae at different soil depths. In control plots fine roots proliferated in upper soil layers, whereas in drought plots there was no fine root growth in upper soil layers and roots mostly occupied deeper soil layers. Fungal hyphae followed the same pattern as fine roots, with the highest biomass in deeper soil layers in drought plots. We conclude that both fine roots and fungal hyphae respond to long-term drought stress by growing into deeper soil layers.

Abstract

Long time series of environmental variables are reflecting the dynamics of ecosystems. Data on climate, water, carbon, nutrients and other observables provide the key to understand terrestrial systems and to detect trends, systemic changes and responses, e.g. to changing climate, disturbances, or management. We present a number of diagnostic measures, based on symbolic dynamics or order statistics, which quantify the information content and the complexity of environmental time series. Three examples for the application of complexity measures in environmental sciences will be provided: Earth System Models and their ability to reproduce observations of Gross Primary Productivity, the dynamics of river runoff, and long-term behavior of ion concentrations in stream water from a monitoring site in Germany. Diagnostic measures of this kind provide data-analytical tools that distinguish different types of natural processes based solely on their dynamics, and are thus highly suitable for environmental science applications such as model structural diagnostics.

Abstract

Convergent Cross Mapping (CCM) has recently been introduced by Sugihara et al. for the identification and quantification of causal relationships among ecosystem variables. In particular, the method allows to decide on the direction of causality; in some cases, the causality might be bidirectional, indicating a network structure. We extend this approach by introducing a method of surrogate data to obtain confidence intervals for CCM results. We then apply this method to time series from stream water chemistry. Specifically, we analyze a set of eight dissolved major ions from three different catchments belonging to the hydrological monitoring system at the Bramke valley in the Harz Mountains, Germany. Our results demonstrate the potentials and limits of CCM as a monitoring instrument in forestry and hydrology or as a tool to identify processes in ecosystem research. While some networks of causally linked ions can be associated with simple physical and chemical processes, other results illustrate peculiarities of the three studied catchments, which are explained in the context of their special history.

To document

Abstract

Data analysis and model-data comparisons in the environmental sciences require diagnostic measures that quantify time series dynamics and structure, and are robust to noise in observational data. This paper investigates the temporal dynamics of environmental time series using measures quantifying their information content and complexity. The measures are used to classify natural processes on one hand, and to compare models with observations on the other. The present analysis focuses on the global carbon cycle as an area of research in which model-data integration and comparisons are key to improving our understanding of natural phenomena. We investigate the dynamics of observed and simulated time series of Gross Primary Productivity (GPP), a key variable in terrestrial ecosystems that quantifies ecosystem carbon uptake. However, the dynamics, patterns and magnitudes of GPP time series, both observed and simulated, vary substantially on different temporal and spatial scales. We demonstrate here that information content and complexity, or Information Theory Quantifiers (ITQ) for short, serve as robust and efficient data-analytical and model benchmarking tools for evaluating the temporal structure and dynamical properties of simulated or observed time series at various spatial scales. At continental scale, we compare GPP time series simulated with two models and an observations-based product. This analysis reveals qualitative differences between model evaluation based on ITQ compared to traditional model performance metrics, indicating that good model performance in terms of absolute or relative error does not imply that the dynamics of the observations is captured well. Furthermore, we show, using an ensemble of site-scale measurements obtained from the FLUXNET archive in the Mediterranean, that model-data or model-model mismatches as indicated by ITQ can be attributed to and interpreted as differences in the temporal structure of the respective ecological time series. At global scale, our understanding of C fluxes relies on the use of consistently applied land models. Here, we use ITQ to evaluate model structure: The measures are largely insensitive to climatic scenarios, land use and atmospheric gas concentrations used to drive them, but clearly separate the structure of 13 different land models taken from the CMIP5 archive and an observations-based product. In conclusion, diagnostic measures of this kind provide dataanalytical tools that distinguish different types of natural processes based solely on their dynamics, and are thus highly suitable for environmental science applications such as model structural diagnostics.

Abstract

Boreal forests contain 30% of the global forest carbon with the majority residing in soils. While challenging to quantify, soil carbon changes comprise a significant, and potentially increasing, part of the terrestrial carbon cycle. Thus, their estimation is important when designing forest-based climate change mitigation strategies and soil carbon change estimates are required for the reporting of greenhouse gas emissions. Organic matter decomposition varies with climate in complex nonlinear ways, rendering data aggregation nontrivial. Here, we explored the effects of temporal and spatial aggregation of climatic and litter input data on regional estimates of soil organic carbon stocks and changes for upland forests. We used the soil carbon and decomposition model Yasso07 with input from the Norwegian National Forest Inventory (11275 plots, 1960–2012). Estimates were produced at three spatial and three temporal scales. Results showed that a national level average soil carbon stock estimate varied by 10% depending on the applied spatial and temporal scale of aggregation. Higher stocks were found when applying plot-level input compared to country-level input and when long-term climate was used as compared to annual or 5-year mean values. A national level estimate for soil carbon change was similar across spatial scales, but was considerably (60–70%) lower when applying annual or 5-year mean climate compared to long-term mean climate reflecting the recent climatic changes in Norway. This was particularly evident for the forest-dominated districts in the southeastern and central parts of Norway and in the far north. We concluded that the sensitivity of model estimates to spatial aggregation will depend on the region of interest. Further, that using long-term climate averages during periods with strong climatic trends results in large differences in soil carbon estimates. The largest differences in this study were observed in central and northern regions with strongly increasing temperatures.

Abstract

We investigate a set of long-term (several decades) time series for the runoff at river gauges at daily resolution. They are monitored by the Agencia Nacional de Aguas, and time series provided by the Operador Nacional do Sistema Elétrico, Brazil. A total of 150 time series was obtained, with an average length of 73 years. Both long-term trends as well as the influence of extreme events on the dynamical behaviour are analyzed. We use Horizontal Visibility Graphs (HVGs) to determine the individual temporal networks for the time series, and extract their degree distributions. Statistical and information-theoretic properties of these distributions are calculated: robust estimators of skewness and kurtosis, the maximum degree occurring in the time series, the Shannon entropy, permutation complexity and Fisher Information. For the latter, we also compare the information measures obtained from the degree distributions to those using the original time series directly, to investigate the impact of graph construction on the dynamical properties as reflected in these measures. Focus is on one hand on universal properties of the HVG, common to all runoff series, and on site-specific aspects on the other. We show that a specific pretreatment of the time series conventional in hydrology, the elimination of seasonality by a separate z-transformation for each calendar day, is highly detrimental to the nonlinear behaviour. It changes long-term correlations and the overall dynamic towards more random behaviour. Analysis based on the transformed data easily leads to spurious results, and bear a high risk of misinterpretation.

To document

Abstract

Measuring energy and matter fluxes between the atmosphere and vegetation using the Eddy Covariance (EC) technique is the state-of-the-art method to quantify carbon exchange between terrestrial ecosystems and their surrounding. The EC equipment is usually mounted onto a flux tower reaching higher than the local canopy. Today, more than 600 flux towers are in operation worldwide. The methodological requirements lead to high sampling frequency (20 Hz) and thus to the production of very long time series. These are related to temperature, wind components, water vapour, heat and gas exchange, and others. In this chapter, the potential of Recurrence Analysis (RA) to investigate the dynamics of this atmosphere-vegetation boundary system is elucidated. In particular, the effect of temporal resolution, the identification of periods particular suitable for reliable EC flux calculations, and the detection of transitions between dynamical regimes will be highlighted.

Abstract

Effects of clear-cut harvesting on ground vegetation plant species diversity and their cover are investigated at two Norway spruce sites in southern Norway, differing in climate and topography. Experimental plots at these two sites were either harvested conventionally (stem-only harvesting) or whole trees including crowns, twigs and branches were removed (whole-tree harvesting), leaving residue piles on the ground for some months. We compare the number of plant species in different groups and their cover sums before and after harvesting, and between the different treatments, using non-parametric statistical tests. An overall loss of ground vegetation biodiversity is induced by harvesting and there is a shift in cover of dominant species, with negative effects for bryophytes and dwarf shrubs and an increase of graminoid cover. Differences between the two harvesting methods at both sites were mainly due to the residue piles assembled during whole-tree harvesting and the physical damage made during the harvesting of residues in these piles. The presence of the residue piles had a clear negative impact on both species numbers and cover. Pile residue harvesting on unfrozen and snow-free soil caused more damage to the forest floor in the steep terrain at the western site compared to the eastern site.

To document

Abstract

Soil organic carbon (C), accumulated over millennia, comprise more than half of the C stored in boreal and temperate forest landscapes. We used the Norwegian national forest inventory and soil survey network (n = 719, no deep organic soils) to explore the validity of a deterministic model representation of this pool (Yasso07). We statistically compared simulated and measured soil C stocks and related differences (measured – simulated) to site factors (drainage, topography, climate, vegetation, C-to-N ratio, and soil classification). Median C stocks were 5.0 kg C·m−2 (model) and 14.5 kg C·m−2 (measurements). Soil C differences related to site factors (r2 of 0.16 to 0.37). For Brunisols, Gleysols, and wet Organic soils, differences related primarily to topographic wetness. For Regosols, Podzols, and Dystric Eluviated Brunisols, they related to climate, profile depth, and, in some cases, drainage class and site index. We argue that soil moisture regimes in our study area overrule tree productivity effects in the determination of soil C stocks and present conditions for soil formation that the model cannot (and does not explicitly) account for. These are processes such as humification and podsolization that involve eluviation and illuviation of dissolved organic C (DOC) with sesquioxides to form spodic B horizons and carbon enrichment due to hampered decomposition in frequently anoxic conditions.

Abstract

Whole-tree harvest (WTH), i.e. harvesting of forest residues (twigs, branches and crown tops) in addition to stems, for bioenergy purposes may lead to biodiversity loss and changes in species composition in forest ground vegetation, which in turn also will affect soil properties. Effects of clear-cut harvesting on ground vegetation have been investigated at two Norway spruce sites in southern east and western Norway, respectively, differing in climate and topography. Experimental plots at these two sites were either harvested conventionally (stem-only harvest, SOH), leaving harvest residues spread on the site,or WTH was carried out, with the residues collected into piles at the site for six - nine months prior to removal. Vegetation plots in the eastern site were established and analysed before WTH and SOH in 2008 and reanalysed after harvesting in 2010, 2012 and 2014. In the western site vegetation plots were established before WTH and SOH in 2010 and reanalysed after harvesting in 2012 and 2014 (and planned for 2016). All vegetation plots are permanently marked. Pre-as well as post-harvesting species abundances of all species in each vegetation plot were each time recorded as percentage cover (vertical projection) and subplot frequency. Environmental variables (topographical, soil physical, soil chemical, and tree variables) were recorded only once; before WTH and SOH. Effec ts of WTH and SOH on ground vegetation biodiversity and cover are presented.

To document

Abstract

The analysis of temporal geospatial data has provided important insights into global vegetation dynamics, particularly the interaction among different variables such as precipitation and vegetation indices. Nevertheless, this analysis is not a straightforward task due to the complex relationships among different systems driving the dynamics of the observed variables. Aiming at automatically extracting information from temporal geospatial data, we propose a new approach to detect stochastic and deterministic patterns embedded into time series and illustrate its effectiveness through an analysis of global geospatial precipitation and vegetation data captured over a 14 year period. By knowing such patterns, we can find similarities in the behavior of different systems even if these systems are characterized by different dynamics. In addition, we developed a novel determinism measure to evaluate the relative contribution of stochastic and deterministic patterns in a time series. Analyses showed that this measure permitted the detection of regions on the global map where the radiation absorbed by the vegetation and the incidence of rain occur with similar patterns of stochasticity. The methods developed in this study are generally applicable to any spatiotemporal data set and may be of particular interest for the analysis of the vast amount of remotely sensed geospatial data currently being collected routinely as part of national and international monitoring programs.

To document

Abstract

The eddy covariance (EC) method is one key method to quantify fluxes in biogeochemical cycles in general, and carbon and energy transport across the vegetation-atmosphere boundary layer in particular. EC data from the worldwide net of flux towers (Fluxnet) have also been used to validate biogeochemical models. The high resolution data are usually obtained at 20 Hz sampling rate but are affected by missing values and other restrictions. In this contribution, we investigate the nonlinear dynamics of EC fluxes using Recurrence Analysis (RA). High resolution data from the site DE-Bay (Waldstein-Weidenbrunnen) and fluxes calculated at half-hourly resolution from eight locations (part of the La Thuile dataset) provide a set of very long time series to analyze. After careful quality assessment and Fluxnet standard gapfilling pretreatment, we calculate properties and indicators of the recurrent structure based both on Recurrence Plots as well as Recurrence Networks. Time series of RA measures obtained from windows moving along the time axis are presented. Their interpretation is guided by three different questions: (1) Is RA able to discern periods where the (atmospheric) conditions are particularly suitable to obtain reliable EC fluxes? (2) Is RA capable to detect dynamical transitions (different behavior) beyond those obvious from visual inspection? (3) Does RA contribute to an understanding of the nonlinear synchronization between EC fluxes and atmospheric parameters, which is crucial for both improving carbon flux models as well for reliable interpolation of gaps? (4) Is RA able to recommend an optimal time resolution for measuring EC data and for analyzing EC fluxes? (5) Is it possible to detect non-trivial periodicities with a global RA? We will demonstrate that the answers to all five questions is affirmative, and that RA provides insights into EC dynamics not easily obtained otherwise.

To document

Abstract

In this chapter, the potential of Recurrence Analysis (RA) for applications in the biogeosciences is demonstrated. We investigate the fraction of absorbed photosynthetically active radiation (FAPAR), an index based on multispectral reflectance properties of land surfaces which relates to the carbon uptake by plants. FAPAR is available with global coverage from satellites. We combine observations from two sensors, SeaWifs on board SeaStar and MERIS on board Envisat, to produce time series with 10 days resolution for a period of 14 years (1998–2011) at a spatial resolution of 0.5∘ latitude × 0.5∘ longitude. After careful quality checking and gap-filling, more than 30,000 individual time series are obtained covering all terrestrial ecosystems and climates apart from Antarctica and major deserts. To characterize the different dynamical behaviour as a function of spatial location, we employ Recurrence Quantification Analysis (RQA) and Recurrence Network Analysis (RNA). They deliver detailed information on the nonlinear dynamics in phase space through embedding. RQA and network measures are calculated for individual time series using identical recurrence parameters, and results are visualized on world maps. Taken together, the recurrence analysis leads to a partitioning of the terrestrial biosphere into regions with distinct dynamical patterns of photosynthetic activity.

To document

Abstract

Tree harvest and different harvesting methods may affect the soil carbon (C) pool in forest ecosystems. In con- ventional stem-only timber harvesting (SOH), branches and tops that are left in the forests may contribute to the build-up of the soil carbon pool. In whole-tree harvesting (WTH), inputs of organic matter from branches and tops are strongly reduced. We established field experiments at Gaupen, SE and Vindberg, SW Norway, to study the short-term effects of SOH and WTH on processes affecting the accumulation and loss of soil C. Logging residues on the WTH plots were collected in piles that were removed after 6 months, rendering two sub treatments (WTH- pile and WTH-removal areas). We weighed selected trees and logging residues, surveyed understorey biomass production, quantified pre-harvest soil C and nutrient pools down to 30 cm. Soil respiration was measured and soil water sampled monthly during the growing season, while temperature and moisture were measured continuously. Organic and mineral horizons were incubated at different temperatures to estimate potential C and N mineraliza- tion, and deep sequencing of the ITS2 barcode region of fungal DNA was performed on the samples. Litterbags were deployed in the SOH plots. The logging residues amounted to 2.2-2.4 kg C m-2 At Gaupen, the mean in situ soil respiration rates increased following harvest with all treatments, but were significantly higher in WTH-pile and SOH relative to the WTH- removal areas in the first year as well as the fourth year of treatment. The former rates included aboveground decomposing needles and twigs but excluded coarser branches. The observed increase in the WTH-removal areas may be related to decomposing roots, as well as to increased C mineralization partly due to the higher soil tem- peratures following harvest. Soil temperature was the single most important factor explaining the variability in soil respiration rates over all treatments. At Vindberg, a decrease in soil respiration was observed with all treatments in the second and third years following harvest. At both sites, decomposition of logging residues from needles was more rapid relative to twigs and fine roots. The decomposing residues released a substantial amount of nitrogen which was gradually reflected in the soil water at 30 cm soil depth. A considerable increase in the NO3-N concen- tration also in the WTH-removal areas in the second year following harvest suggests an increase in N availability from decomposing fine roots and/or soil organic matter. The increased N availability in the WTH-removal areas was supported by results from short term lab incubations of undisturbed soil from the forest floor. The changes in the WTH-removal areas were also reflected in the soil fungal diversity: saprophytic ascomycetes on decaying plant material showed a striking increase in all treatments. For the WTH-removal areas, this may, again, be related to the increased input of root litter; however, the decrease in mycorrhizal basidiomycete species and the vigorous increase of ascomycetes following harvest may also affect the C mineralization of soil organic matter.

Abstract

In an attempt to discern stochastic and deterministic parts of measured signals, we analyze time series from the viewpoint of ordinal pattern statistics. After choosing a suitable embedding dimension $D$, the occurrencies of all $D!$ patterns form a probability distribution $P$. The latter is input to information and complexity functionals describing, e.g., chaotic regimes or stochastic properties due to long-range correlations. Here, we use an information quantifier which is local in pattern probability space, the Fisher information $F$. This is calculable only after fixing a pattern coding scheme, i.e. numbering each and every pattern. It has been demonstrated that $F$ discerns different dynamic regimes for the logistic map to a certain extent; however, this depends on the details of the coding scheme. Here, we seek to find an optimal coding scheme for long-range correlated stochastic processes, mimicking many records e.g. from the geosciences. To increase the contrast between colored noise and deterministic processes, $F$ should be minimal for the former. Structurally similar ordinal patterns should be located adjacent to each other. Similarity is related to the number of inversions in the respective patterns. In practical terms, it is impossible to try all $D!!$ coding schemes whenever$D > 3$; however, we demonstrate a classification of coding schemes into equivalence classes based on the number of "jumps" in the patterns. These are used to improve the Keller and Lehmer coding schemes. The approach has a potential to provide an analytical understanding of the Fisher information for stochastic processes. Results for these optimizations will be shown for both the logistic map and colored ($k$-) noise. As a byproduct, an innovative method to estimate the scaling exponent $k$ emerges. Finally, we comment shortly on the importance of finite size effects, which is always an issue when dealing with observed data.

Abstract

Reliable methods are required to predict changes in soil carbon stocks. Process-based models often require many parameters which are largely unconstrained by observations. This induces uncertainties which are best met by using repeated measurements from the same sites. Here, we compare two carbon models, Yasso07 and Romul, in their ability to reproduce a set of field observations in Norway. The models are different in the level of process representation, structure, initialization requirements and calibration- and parameterization strategy. Field sites represent contrasting tree species, mixture and soil types. The number of repetitions of C measurements varies from 2 to 6 over a period of up to 35 years, and for some of the sites, which are part of long-term monitoring programs, plenty of auxiliary information is available. These reduce the danger of overparametrization and provide a stringent testbed for the two models. Focus is on the model intercomparison, using identical site descriptions to the extent possible, but another important aspect is the upscaling of model results to the regional or national scale, utilizing the Norwegian forest inventory system. We suggest that a proper uncertainty assessment of soil C stocks and changes has to include at least two (and preferably more) parametrized models.

Abstract

Reliable methods are required to predict changes in soil carbon stocks. Process-based models often require many parameters which are largely unconstrained by observations. This induces uncertainties which are best met by using repeated measurements from the same sites. Here, we compare two carbon models, Yasso07 and Romul, in their ability to reproduce a set of field observations in Norway. The models are different in the level of process representation, structure, initialization requirements and calibration- and parameterization strategy. Field sites represent contrasting tree species, mixture and soil types. The number of repetitions of C measurements varies from 2 to 6 over a period of up to 35 years, and for some of the sites, which are part of long-term monitoring programs, plenty of auxiliary information is available. These reduce the danger of overparametrization and provide a stringent testbed for the two models. Focus is on the model intercomparison, using identical site descriptions to the extent possible, but another important aspect is the upscaling of model results to the regional or national scale, utilizing the Norwegian forest inventory system. We suggest that a proper uncertainty assessment of soil C stocks and changes has to include at least two (and preferably more) parametrized models.

Abstract

A wide range of forest products and industries have been examined in life cycle analyses (LCA). Life cycle data are essential for identifying forestry operations that contribute most to carbon emissions. Forestry can affect net CO2 emissions by changing carbon stocks in biomass, soil and products, by supplying biofuels to replace fossil fuels as well as by establishing new forests. The transport of forest products is crucial to greenhouse gas (GHG) emissions. We conceptualize the chain from seed production, silviculture, harvesting, and timber transport to the industry as a system. Inputs to the system are energy and fuel, the output represents GHG emissions. The reference functional unit used for the inventory analysis and impact assessment is one cubic meter of harvested timber under bark. GHG emissions from forestry in East Norway were calculated for the production of one such unit delivered to the industry gate in 2010 (cradle-to-gate inventory), showing that timber transport from the forest to the final consumer contributed with more than 50 % to the total GHG emissions. To assess uncertainty of model approaches, the LCA was conducted with two different models, SimaPro and GaBi, both using the Ecoinvent database with data adapted to European conditions.

Abstract

A wide range of forest products and industries have been examined in life cycle analyses (LCA). Life cycle data are essential for identifying forestry operations that contribute most to carbon emissions. Forestry can affect net CO2 emissions by changing carbon stocks in biomass, soil and products, by supplying biofuels to replace fossil fuels as well as by establishing new forests. The transport of forest products is crucial to greenhouse gas (GHG) emissions. We conceptualize the chain from seed production, silviculture, harvesting, and timber transport to the industry as a system. Inputs to the system are energy and fuel, the output represents GHG emissions. The reference functional unit used for the inventory analysis and impact assessment is one cubic meter of harvested timber under bark. GHG emissions from forestry in East Norway were calculated for the production of one such unit delivered to the industry gate in 2010 (cradle-to-gate inventory), showing that timber transport from the forest to the final consumer contributed with more than 50 % to the total GHG emissions. To assess uncertainty of model approaches, the LCA was conducted with two different models, SimaPro and GaBi, both using the Ecoinvent database with data adapted to European conditions.

To document

Abstract

The ingrowth core method is widely used to assess fine root (diameter < 2 mm) production but has many inherent deficiencies. In this study, we modified this method by adopting mini ingrowth cores (diameter 1.2 cm), extending sample intervals to a growing season, and developing new models to quantify the concurrent production, mortality and decomposition, and applied them to a secondary Mongolian oak (Quercus mongolica Fischer ex Ledebour) forest. Annual fine root production, mortality and decomposition estimated by our method were 2.10 ± 0.23, 1.78 ± 0.20 and 0.85 ± 0.13 t ha−1, respectively, and 33.3% of the production was decomposed in the growing season. The production estimate using our method was significantly higher than those using two long-term ingrowth core (sample interval >2 months) methods. However, it was significantly lower than that using the short-term ingrowth core (sample interval <2 months) method, presumably due to the lower root competition and less decomposition occurring in the short-term cores. The fine root estimates using our method in the growing season were generally higher than those using the forward and continuous inflow methods but lower than those using the backward method. Our method reduces the disturbances in roots and soil, minimizes the sampling frequency and improves the quantification of fine root decomposition during the sample intervals. These modifications overcome the limitations associated with the previous ingrowth core methods. Our method provides an improved alternative for estimating fine root production, mortality and decomposition.

Abstract

Over the past 40 years, a new multidisciplinary field of study has emerged which is characterised by at least two major changes in the way some scientists treat systems. First, it is increasingly accepted that we cannot fully understand the laws that govern a system simply by studying its parts, nor can we fully understand the behaviour of the parts without placing them in the context of the larger system in which they are embedded. This realization, which has arisen as we face the limits of reductionist science, has given rise to the development of new models and methods that facilitate the study of systems across multiple scales of organization. Second, the notions of equilibrium and predictability in natural systems, developed in the 19th Century and continuously pursued until far into the 20th Century, are being rejected in favour of models that embrace variability, diversity, continual change and adaptation as the status quo. Traditional analytical models that assume a stable equilibrium are being replaced by new approaches that facilitate the exploration of a system’s natural range of variation and its possible emergent responses to changing external conditions. The implications of this new field, now known as complexity science, are manifest across disciplines, fundamentally changing the way we study, analyze and perceive natural systems. We provide an overview of complexity science in the context of forest management.

To document

Abstract

The quality of surface water and groundwater is closely related to flow paths in the vadose zone. Therefore, dye tracer studies are often carried out to visualise flow patterns in soils. These experiments provide images of stained soil profiles and their evaluation demands knowledge in hydrology as well as in image analysis and statistics. The classical analysis consists of image classification in stained and non-stained parts and calculation of the dye coverage (i.e. the proportion of staining). The variation of this quantity with depth is interpreted to identify dominant flow types. While some feature extraction from images of dye-stained profiles is necessary, restricting the analysis to the dye coverage alone might miss important information. In our study we propose to use several index functions to extract different (ideally complementary) features. We associate each image row with a feature vector (i.e. a certain number of image function values) and use these features to cluster the image rows to identify similar image areas. Because images of stained profiles might have different reasonable clusterings, we calculate multiple consensus clusterings. Experts can explore these different solutions and base their interpretation of predominant flow type on quantitative (objective) criteria.

To document

Abstract

Fine roots contribute to net primary production in forests, but knowledge of fine root longevity and turnover is still incomplete and limited to few tree species. In this study, we used minirhizotrons to compare fine root biomass, longevity and turnover of Pinus sylvestris L., Betula pendula Roth and Picea abies (L) Karst. in southern Sweden. Minirhizotron tubes were installed in 2006 and root images were taken in 2007–2010. Soil cores were used to estimate fine root biomass. Soil samples were taken from the humus layer and from 0 to 10 cm, 10 to 20 cm and 20 to 30 cm depth in the mineral soil. Only images from the humus layer and the upper 10 cm of mineral soil were included in root analysis. Spruce has a higher aboveground production than pine and birch in southern Sweden and this was reflected in larger fine root biomass as well as higher fine root biomass production. The annual tree fine root biomass production (humus and 0–30 cm in mineral soil) was 73, 78 and 284 g m−2 in pine, birch and spruce stands, respectively. Thicker fine roots tended to live longer. The majority of the fine roots were thinner than 0.5 mm in diameter, with a turnover rate (KM) of 0.4 year−1. When comparing all fine roots, i.e. all roots 0–2 mm, pine had the highest longevity, 1120 days, compared with 900 days for spruce and 922 days for birch (KM).

To document

Abstract

When calculating the Bandt and Pompe ordinal pattern distribution from given time series at depth D, some of the D! patterns might not appear. This could be a pure finite size effect (missing patterns) or due to dynamical properties of the observed system (forbidden patterns). For pure noise, no forbidden patterns occur, contrary to deterministic chaotic maps. We investigate long time series of river runoff for missing patterns and calculate two global properties of their pattern distributions: the Permutation Entropy and the Permutation Statistical Complexity. This is compared to purely stochastic but long-range correlated processes, the k-noise (noise with power spectrum f−k), where k is a parameter determining the strength of the correlations. Although these processes closely resemble runoff series in their correlation behavior, the ordinal pattern statistics reveals qualitative differences, which can be phrased in terms of missing patterns behavior or the temporal asymmetry of the observed series. For the latter, an index is developed in the paper, which may be used to quantify the asymmetry of natural processes as opposed to artificially generated data.

Abstract

Groundwater pollution by agrochemicals, degradation of soil quality and pollution of aquatic ecosystems by agricultural drainage waters have become an issue in the last decades. Flow processes in the vadose zone are closely related to these problems. In general, water flow in soils can be classified into two major categories: uniform and non-uniform (preferential) flow (In: U.S. National Committee for Rock Mechanics, Conceptual Models of Flow and Transport in the Fractured Vadose Zone, 2001, pp.149-187). The former describes a relatively slow movement of water through the porous soil matrix and can be modelled by Richard”s equation. The latter comprises all flow types where water bypasses a portion of the soil matrix and flows through localised (i.e. preferential) paths. Unlike uniform flow, preferential flow is hardly predictable because the assumptions of Richard”s equation of a homogeneous representative elementary volume characterised by a single value of water potential, water content and hydraulic conductivity are frequently violated (Eur J Soil Sci, 2007; 58:523-546)....

Abstract

A crucial process of the terrestrial carbon cycle is photosynthetic uptake through plants. This may be quantified by calculating the fraction of absorbed photosynthetic active radiation (fapar), based on multispectral reflectance properties of the earth surface. The fapar index is available with global coverage from satellite sensors.Here, we combine two satellite missions, SeaWifs on board OrbView2 and MERIS on board Envisat, to produce time series with 10 days resolution for a period of 14 years (19982011) at a spatial resolution of 0.5 latitude x 0.5 longitude. These more than 50000 individual time series represent a huge range of dynamical behavior with respect to variability, periodicities and correlation structure.To characterize differences as a function of spatial location or distance, we employ Recurrence Quantification Analysis (RQA) and Recurrence Network Analysis (RNA). Two strategies are followed. On one hand, RQA and network variables are calculated for individual time series using identical recurrence parameters, and compared to see whether differences between them resemble different climate regimes, biomes, plant functional types or landuse classes. On the other hand, a multivariate extension of RNA will be exploited to see whether networks within networks occur, i.e. whether RNA provides sufficient contrasts to discern different clusters of pixels on the globe.Taken together, the recurrence analysis might lead to a new classification of the terrestrial biosphere which in turn can be compared to existing partitioning based on climate and/or vegetation properties. A number of technical issues will be addressed as well, such as the impact of the finite length of the series (504 values each), the necessity to gapfill parts of the data, the stability of network variables against changes in the recurrence parameters, or the computational challenges involved in the multinetwork analysis of many series. http://dames.pik-potsdam.de/Abstracts.pdf

Abstract

Our aim is to investigate the temporal dynamics of the Fraction of Absorbed Photosynthetically Active Radiation (fAPAR) on a global scale and its relation to the main meteorological variables across space. We focus on complex patterns in time, which are neither regular (trend and seasonality) nor random (noise), but somewhere in between. We quantify complexity and information content or entropy using methods from order statistics and complexity sciences.Time series with high entropy are difficult to predict, whereas time series with high complexity are difficult to describe. This leads to a spatially explicit characterization of complex patterns in a very sensitive way. We use FAPAR observations (SeaWiFS and MERIS, 1998 to 2012) along with gridded global surface air temperature, precipitation and shortwave radiation.All these time series are explored on a pixelbypixel basis and clustered according to a very recent classification system of the land surface. In addition, we quantify the time reversal asymmetry of these data. We compare environmental time series with data from a stochastic candidate process temporally symmetric and long range correlated artificial knoise.Results were plotted in the ComplexityversusEntropy plane (CH plane), showing the particular footprint of each variable in a very sensitive way. Visualized in world maps, results revealed unexpected complex pattern in some dry regions, in particular on pixels surrounding deserts and in eastern Sahara. In this respect, the results provide a new classification of the climate and the biosphere. http://dames.pik-potsdam.de/Abstracts.pdf

Abstract

We exploit two recently developed informationtheoretic quantities (or measures) designed to quantify information content and complexity of ordered data (time series), respectively. Both are based on order statistics of given data sets, and probe into the shortterm structure of the data only due to finite length restrictions. Their usage requires fixing one parameter, word length or order depth D. The information measure is the orderbased Shannon entropy HS, and the complexity measures is the JensenShannon divergence CJS. The latter requires a chosen reference distribution, i.e. CJS represents a class of measures. Entropy HS and complexity CJS of data series may be represented against each other in a twodimensional diagram which we will refer to as ComplexityEntropy Causality Plane, or CECP. Very long realizations of classic stochastic processes and chaotic deterministic maps each obey one location in the CECP, specific for the process. This can be used to differentiate chaos from correlated noise (Rosso et al. 2007), which is notoriously difficult otherwise. For observed data, a mixture of deterministic (signal) and stochastic (noise) parts is to be expected. We use an ensemble of longterm river runoff time series as example, which are known to exhibit powerlaw decaying longrange correlations. We compare these data with a longrange correlated candidate process, the k noise, from the perspective of order statistics and the CECP. Although these processes resemble runoff series in their correlation behavior and may be even tuned to any runoff series by changing the value of k, the CECP locations and in particular the order pattern statistics reveals qualitative differences between them. We give a detailed account of these differences, and use them to conclude on the deterministic nature of the (shortterm) dynamics of the runoff time series. The proposed methodology also represents a stringent test bed for hydrological or other environmental models. http://dames.pik-potsdam.de/Abstracts.pdf

Abstract

We calculate entropy and complexity of runoff time series and artificially generated series with long-range correlations. Entropy and complexity of data series may be represented against each other in a two-dimensional diagram which we will refer to as Complexity-Entropy Causality Plane, or CECP. We use a recently developed framework for these two indicators based on order statistics. It is well-known that runoff, as all other environmental time series actually measured, is a mixture of deterministic (signal) and stochastic (noise) parts, the latter due to noise inherent in the measurement process and externally induced by natural processes. The distinction between signal and noise is notoriously difficult and subject to much debate. In our approach, the observed series are compared to purely stochastic but long-range correlated processes, the k noise, where k is a parameter determining the strength of the correlations. Although these processes resemble runoff series in their correlation behavior and may be even tuned to any runoff series by changing the value of k, the CECP locations and in particular the order pattern statistics reveals qualitative differences between runoff and k noise. We use these differences to conclude on the deterministic nature of the (short-term) dynamics of the runoff time series. The proposed methodology also represents a stringent test bed for hydrological models.

Abstract

Spatial dependencies among environmental variables are often quantified by spatial autocorrelation functions. The latter basically assume linearity and isotropy, requirements usually not satisfied for measured data. Typical symptoms of violated assumptions are biased parameter estimations. Relaxing the assumptions of linear dependencies and isotropy, we present a conceptual generalization of spatial analysis where locally varying anisotropies in the geographical space are uncovered by investigating nonlinear dependencies among observations. The framework is illustrated by generalizing two examples: distance decay relations and spatial filtering. The proposed alternative is based on geodesic ecological and anisotropic spatial distances.

Abstract

Long-term monitoring of headwater semi-natural catchments is used to document persistence and changes in ecosystems. At three headwater catchments in the Bramke basin in Northern Germany, physical and chemical variables in rainfall, soil solution from various depths (20–300 cm) and streamwater have been monitored. The Lange Bramke catchment is largely covered by a Norway spruce (Picea abies, Karst.) stand planted in the 1950ies. Over 29 years, 4310 water samples from streamwater and 5475 soil water samples were analysed for major constituents. Both linear methods (principal component analysis (PCA) and cross correlation (CC)) as well as non-linear methods (isometric feature mapping (ISOMAP) and maximum variance unfolding (MVU)) were used to analyze the spatiotemporal patterns of dissolved major ion concentrations in soil solution and streamwater. This approach provides a multiscale characterisation of links between soil water and streamwater at the catchment scale. Pattern identification augments the interpretation of processes in terms of transport and storage. The long time scales were dominated by trends in ions implicated in soil acidification. This reflects the decreasing input of acid deposition. At the annual scale, where hydrological effects dominate, each of the three adjacent catchments showed different patterns. Various empirical and process-based models have been applied in the past to the Bramke catchments. Results of the data-oriented approach can be used to indicate the potential and limits of process-oriented models for this data set.

Abstract

Remote sensing of the activity of vegetation in relation to environmental conditions provides an invaluable basis for investigating the spatiotemporal dynamics and patterns of variability for ecosystem processes. We investigate the fraction of Absorbed Photosynthetically Active Radiation (fAPAR) using SeaWiFS satellite observations from 1998 to 2005 and ancillary meteorological variables from the CRU-PIK dataset with a global coverage at a spatial resolution of 0.5o x 0.5o. A pixel-by-pixel spectral decomposition using Singular System Analysis leads to a global “classification” of the terrestrial biosphere according to prevalent time-scale dependent dynamics of fAPAR and its relation to meteorology. A complexity analysis and a combined subsignal extraction and dimensionality reduction reveals a series of dominant geographical gradients, separately for different time scales. At the annual scale, which explains around 50% of the fAPAR variability as a global average, patterns largely resemble the biomes of the world as mapped by biogeographical methods, and are driven by temperature and by pronounced rain seasons in the tropics. On shorter time scales, fAPAR fluctuations are exclusively driven by water supply, inducing, e.g., semiannual cycles in the equatorial belt of Africa or the Indo-Gangetic Plain. For some regions however, in particular South America, altitude, mean temperature, drought probability and fire occurrences are parameters that seem to shape the spatial patterns of fAPAR across time scales. Overall, we provide a first global multiscale characterization of fAPAR and highlight different mechanisms in land-surface-climate couplings.

Abstract

There is a great demand for involving rapid, non destructive and less time consuming methods for quick control and prediction of soil quality, environmental monitoring, and other precision measurements in agriculture. Near infrared reflectance spectroscopy (NIRS) is considered as an appropriate alternative method to conventional analytical methods for large scale measurements. The objective of this study was to investigate the possibilities of NIRS for prediction of some chemical properties of soil samples. A total of 97 samples from Stara Zagora, Kazanlak and Gurkovo region taken from the 0-40 cm layer were collected. Soil types were Luvisols, Vertisols, Fluvisols and Rankers. The samples were analyzed for total phosphorus by spectrometric determination of phosphorus soluble in sodium hydrogen carbonate solution, total nitrogen by Kjeldahl method, pH (H O)-potentiometrically and electrical conductivity (EC). 2 The spectral data of all air dried samples were measured using an Perkin Elmer Spectrum One NTS, FT-NIR Spectrometer, within the range from 700 to 2500 nm. Partial Least Squares (PLS) regression was used to built models to determine soil chemical parameters from the NIR spectra. Two-third of the samples were used as a calibration set and the remaining samples as independent validation test set. The best model was obtained for total nitrogen with a coefficient of determination r=0,91, standard error of calibration SEP=336 mg/kg, and the ratio of the standard variation of the reference data to the SEP, indicating the performance of the calibration, of RPD=2,3. The accuracy of prediction was poor for electrical conductivity. The results obtained clearly indicated that NIRS had the potential to predict some soil components rapidly and without sample preparation.

Abstract

A changing climate will likely influence the selection of tree species in the future, and this may in turn affect the size of the pools and fluxes of carbon. Tree species differ in growth rate, fine-root turnover and quality of litter and tend to produce different types of understory vegetation. In Sweden three tree species (Norway spruce [Picea abies] 43%, Scots pine [Pinus sylvestris] 39% and birch [Betula spp.] 11%) dominate. In the present study we used field experiments in southern Sweden to test if these tree species differed in root distribution and turnover.

Abstract

For tracer studies at the catchment scale, travel times are often assumed to be stationary. We question the validity of this assumption. We analyzed a series of tracer experiments conducted under exceptionally controlled conditions at Gårdsjön, Sweden. The Gårdsjön G1 catchment was covered by a roof underneath which natural throughfall has been replaced by artificial irrigation with a pre-defined chemical composition. This unique setup was used to perform replicated catchment scale Br tracer experiments under steady state storm flow conditions in five different years. A log-normal distribution function was fitted to all Br breakthrough curves. Fitted parameter values differed significantly for some of the experiments. These differences were not only related to the slightly different hydrologic boundary and initial conditions for the experiments, but also to seasonal changes in catchment properties that may explain the different flow paths during the experiments. We conclude that the travel time distribution is not only linked to discharge but also explicitly related to other water fluxes such as evapotranspiration, and that it is not stationary even under steady-state flow conditions. Since the attenuation of soluble pollutants is fundamentally linked to the travel times of water through the subsurface of a catchment, it is of crucial importance to understand the latter in detail. However, it is still unclear which are the dominant processes controlling their distribution.

To document

Abstract

The Eurasian spruce bark beetle, Ips typographus, is one of the major forest insect pests in Europe, capable of mass-attacking and killing mature Norway spruce trees. The initiation and development of a new generation are strongly controlled by temperature and a warmer climate may affect the number of generations that is produced per year and hence the outbreak dynamics. Experimental knowledge regarding reproductive diapause adaptations is, however, too sparse for largescale assessments of future trends. We developed a model description of diapause induction, and used gridded observational temperature data to evaluate multiple combinations of day length and temperature thresholds to find the model parameterisation most coherent with I. typographus monitoring data from Scandinavia. The selected model parameterisation is supported by European literature data, though further experimental studies are required to analyse population specific adaptations and capacity for adjustments to changing climate conditions. Implementing the model description of reproductive diapause in a temperature driven model of bark beetle phenology (swarming activity and development from egg to mature bark beetle), enabled us to assess the length of the late summer swarming period that is a critical determinant of the risk of forest damage. By using regional climate model data we show that higher temperatures can result in increased frequency and length of late summer swarming events, producing a second generation in southern Scandinavia and a third generation in lowland parts of central Europe. Reproductive diapause will not prevent the occurrence of an additional generation per year, but the day length cues may restrict the length of the late summer swarming period.

Abstract

Fungi are the main degraders of organic matter and are associated symbiotically with over 80% of terrestrial plants (Smith and Read 1997). Thus, the extent of the mycelial network is an indicator of the decomposing or symbiotic activity. Although the importance of fungi in soil is undisputable, the determination of the extent of hyphal mats and the hyphal biomass is difficult to assess. Methods for estimating hyphae in soil are mostly based on the gridline intersect method originally developed to determine the root length or recently by measuring of the ergosterol content, fungal sterol found in the cell membranes....

Abstract

The release of carbon dioxide (CO2) from the land surface via different respiratory processes is a major flux in the global carbon cycle, antipodal to CO2 uptake via photosynthesis. Understanding the sensitivity of respiratory processes to temperature is central for quantifying the climate–carbon cycle feedback. In a recent study we approximated the sensitivity of terrestrial ecosystem respiration to air temperature (Q10) across 60 FLUXNET sites. For this objective, we developed a novel methodology that circumvents seasonally confounding effects. Contrary to previous findings, our results suggest that Q10 is independent of mean annual temperature, does not differ among biomes, and is confined to values around 1.4 ± 0.1. However, the shape of the strong relation between photosynthesis and respiration is highly variable among sites. The results may partly explain a less pronounced climate–carbon cycle feedback than suggested by current carbon cycle climate models. In the talk we put our findings into context with other recent results and critically discuss their potential for evaluating temperature sensitivity of respiration in terrestrial biosphere models and parameterizing future land surface schemes.

Abstract

In ecosystem research, data-driven approaches to modeling are of major importance. Models are more often than not shaped by the spatiotemporal structure of the observations: an inverse modeling approach prevails. Here, I investigate the insights obtained from Recurrence Quantification Analysis of observed ecosystem time series. As a typical example of available long-term monitoring data, I choose time series from hydrology and hydrochemistry. Besides providing insights into the nonstationary and nonlinear dynamics of these variables, RQA also enables a detailed and temporally local model-data comparison.

Abstract

We estimated the sensitivity of terrestrial ecosystem respiration to air temperature across 60 FLUXNET sites by minimizing the effect of seasonally confounding factors. Graf et al. now offer a theoretical perspective for an extension of our methodology. However, their critique does not change our main findings and, given the currently available observational techniques, may even impede a comparison across ecosystems.

Abstract

The airborne laser scanning (ALS) penetration rate, i.e. the ratio of ground echoes to total echoes, is a proxy for gap fraction. Hence, ALS has a potential for monitoring forest properties that are related to gap fraction, such as leaf area index, canopy cover and disturbance. Furthermore, two gap types may be distinguished: While a pulse that only produces a ground echo most likely hit a large, between-tree gap, a pulse that produces a ground echo as the last of several returns most likely hit smaller, within-canopy gaps. This may be utilized to distinguish between disturbance types such as defoliation and tree removal. However, the ALS penetration rate needs to be calibrated with gap fraction measurements in the field, because it is influenced by technical properties of the acquisition. The aim of this study was to quantify the magnitude of this influence, by comparing repeated acquisitions with different technical specifications. We had at hand 12 ALS acquisitions which could be combined into six pairs, from four spruce and pine dominated forests in Norway. We established 20x20 m grids, and for each grid cell we extracted three penetration variables: first echo penetration, last-of-many echo penetration, and total (i.e., first and last echo). We log-transformed the penetration variables (P1 and P2) from two laser acquisitions, and fitted the no-intercept, linear model log(P1) = log(P2), applying total least squares regression analysis. In a majority of the cases, the penetration variables were very similar, i.e. they deviated by <10%. For the first echo penetration the slopes varied from 0.87 to 1.07 and the R2 values ranged between 0.91 and 0.99. For the last-of-many echo penetration, there was generally weaker correspondence with slopes varying from 0.78 to 1.02, and R2 values ranging from 0.60 to 0.94. Finally, for the total penetration there was again stronger agreement with slopes in the range 0.83-1.03 and R2 values from 0.88 to 0.99. In conclusion, it seems that the penetration ability of different ALS scans in many cases are very similar, and further research may reveal ranges of standardized settings for which field inventory can be redundant.

To document

Abstract

Terrestrial biosphere models are indispensable tools for analyzing the biosphere-atmosphere exchange of carbon and water. Evaluation of these models using site level observations scrutinizes our current understanding of biospheric responses to meteorological variables. Here we propose a novel model-data comparison strategy considering that CO2 and H2O exchanges fluctuate on a wide range of timescales. Decomposing simulated and observed time series into subsignals allows to quantify model performance as a function of frequency, and to localize model-data disagreement in time. This approach is illustrated using site level predictions from two models of different complexity, Organizing Carbon and Hydrology in Dynamic Ecosystems (ORCHIDEE) and Lund-Potsdam-Jena (LPJ), at four eddy covariance towers in different climates. Frequency-dependent errors reveal substantial model-data disagreement in seasonal-annual and high-frequency net CO2 fluxes. By localizing these errors in time we can trace these back, for example, to overestimations of seasonal-annual periodicities of ecosystem respiration during spring greenup and autumn in both models. In the same frequencies, systematic misrepresentations of CO2 uptake severely affect the performance of LPJ, which is a consequence of the parsimonious representation of phenology. ORCHIDEE shows pronounced model-data disagreements in the high-frequency fluctuations of evapotranspiration across the four sites. We highlight the advantages that our novel methodology offers for a rigorous model evaluation compared to classical model evaluation approaches. We propose that ongoing model development will benefit from considering model-data (dis)agreements in the time-frequency domain.

Abstract

Today the spruce bark beetle Ips typographus is always univoltine in Northern Europe including Norway and completes development from egg to adult between May and August. Further south in Europe, development is bivoltine with the completion of two generations in most years. A temperature-driven developmental model suggests that by 2070-2100 the voltinism of I. typographus will change dramatically in Norway. If summers become only 2.5°C warmer than today bivoltinism can be expected every single year in the major spruce growing areas in S-Norway. This is likely to have dramatic effects on forestry since two generations per year will give two, instead of one, attack periods each summer. In addition to increasing the number of attacked trees the effect of the attacks may also be more severe, as Norway spruce is more susceptible to beetle attacks later in the summer. However, climate change will probably also change the phenology of Norway spruce and thus its susceptibility to attack by I. typographus and its phytopathogenic fungal associates. We are currently modelling how tree resistance varies with temperature and tree phenology in order to provide more well-founded advice to forest managers on the interaction between bark beetles and tree in a future climate.

To document

Abstract

The respiratory release of carbon dioxide (CO2) from the land surface is a major flux in the global carbon cycle, antipodal to photosynthetic CO2 uptake. Understanding the sensitivity of respiratory processes to temperature is central for quantifying the climate–carbon cycle feedback. We approximated the sensitivity of terrestrial ecosystem respiration to air temperature (Q10) across 60 FLUXNET sites with the use of a methodology that circumvents confounding effects. Contrary to previous findings, our results suggest that Q10 is independent of mean annual temperature, does not differ among biomes, and is confined to values around 1.4 ± 0.1. The strong relation between photosynthesis and respiration, by contrast, is highly variable among sites. The results may partly explain a less pronounced climate–carbon cycle feedback than suggested by current carbon cycle climate models.

Abstract

Understanding the feedback between terrestrial biosphere processes and meteorological drivers is crucial to ecosystem research as well as management. For example, remote sensing of the activity of vegetation in relation to environmental conditions provides an invaluable basis for investigating the spatiotemporal dynamics and patterns of variability. We investigate the Fraction of Absorbed Photosynthetically Active Radiation (fAPAR) using SeaWiFS satellite observations from 1998 to 2005 and ancillary meteorological variables from the CRU-PIK dataset. To what extent do precipitation and temperature dominate the terrestrial photosynthetic activity on monthly to interannual time scales? A spectral decomposition using Singular System Analysis leads to a global ‘classification’ of the terrestrial biosphere according to prevalent time-scale dependent dynamics of fAPAR and its relation to the meteorology. A complexity analysis and a combined subsignal extraction and dimensionality reduction reveals a series of dominant geographical gradients, separately for different time scales. Here, we differentiate between three time scales: on short time scales (compared to the annual cycle), variations in fAPAR coincide with corresponding precipitation dynamics. At the annual scale, which explains around 50% of the fAPAR variability as a global average, patterns largely resemble the biomes of the world as mapped by biogeographic methods.At longer time scales, spatially coherent patterns emerge which are induced by precipitation and temperature fluctuations combined. However, we can also identify regions where the variability of fAPAR on specific time scales cannot be traced back to climate and is apparently shaped by other geoecological or anthropogenic drivers. http://uregina.ca/prairies/assets/Prairie_Summit_Final_Program.pdf

Abstract

Determining the feedbacks between terrestrial biosphere processes and the meteorological drivers (here precipitation and temperature) is crucial to ecosystem research. In this context, the continuous monitoring of the earth surface provides an invaluable basis for investigating the spatiotemporal dynamics in the activity of vegetation in relation to environmental conditions. Here, we seek to identify which patterns of variability in the meteorological drivers dominate the terrestrial photosynthetic activity from monthly to interannual time scales (resp. fluctuation frequencies). We investigate the Fraction of Absorbed Photosynthetically Active Radiation (FAPAR) using SeaWiFS observations from 1998 to 2005 and ancillary meteorological variables. A spectralanalysis leads to a global `classification` of the terrestrial biosphere according to prevalent scale dependent dynamics of fAPAR and its relation to the meteorology. A combined subsignal extraction and dimensionality reduction reveals a series of dominant geographical gradients on specific time scales. E.g. we uncover spatially coherent patterns at low frequencies and show where these are induced by precipitation or temperature fluctuations. We also show where high frequency variations (relative to the annual cycle) in fAPAR coincide with corresponding precipitation dynamics. However, we can also identify regions where the variability of fAPAR on specific time scales cannot be traced back to climate and is apparently shaped by other geoecological or anthropogenic drivers. http://www.terrabites.net/fileadmin/user_upload/terrabites/PDFs/Programme_Book_TERRABITES.pdf

To document

Abstract

Information retrieval from spatiotemporal data cubes is key to earth system sciences. Respective analyses need to consider two fundamental issues: First, natural phenomena fluctuate on different time scales. Second, these characteristic temporal patterns induce multiple geographical gradients. Here we propose an integrated approach of subsignal extraction and dimensionality reduction to extract geographical gradients on multiple time scales. The approach is exemplified using global remote sensing estimates of photosynthetic activity. A wide range of partly well interpretable gradients is retrieved. For instance, well known climate-induced anomalies in FAPAR over Africa and South America during the last severe ENSO event are identified. Also, the precise geographical patterns of the annual–seasonal cycle and its phasing are isolated. Other features lead to new questions on the underlying environmental dynamics. Our method can provide benchmarks for comparisons of data cubes, model runs, and thus be used as a basis for sophisticated model performance evaluations.

Abstract

The utilization history of the Lange Bramke catchment and the northern Harz mountains is dominated by ore mining. Historical documents were used to provide ample evidence that forestry and water utilization were managed according to administrative goals in a largely centralized manner. However, the perception of the landscape and its function and purpose have changed significantly over the centuries. In particular, the distinction between renewable (such as forests) and non-renewable resources (such as ore deposits) is a rather modern one, as is the principle of sustainability. This change in perception is apparent from the type of maps used, the different conflicts on property and exploitation rights, and the request for quantitative inventories of resources, appearing only quite late in the mining history. The remnants of smelters and charcoal production still demonstrate the importance of historical land use for proper interpretation of monitoring data.

Abstract

In European forests, standing stocks are currently higher than ever during the last decades, in part due to reduced logging or the abandonment of agricultural land. However, data from intensive monitoring plots reveal an increased growth even without direct human intervention.We used a set of 363 plots from 16 European countries to investigate the influence of environmental factors on forest growth: nitrogen, sulphur and acid deposition, temperature, precipitation and drought, for Norway spruce, Scots pine, common beech and European as well as sessile oak.We used existing information on site productivity, stand age and stand density to estimate expected growth. Relative tree growth, i.e., the ratio between actual growth within a five-year period and expected growth, was then related to environmental factors in a stepwise multiple regression.The results consistently indicate a fertilizing effect from nitrogen deposition, with roughly one percent increase in site productivity per kg of nitrogen deposition per ha and year, or 20 kg C fixation per kg N deposition. This was most pronounced for plots having soil C/N ratios above 25. We also found a positive albeit less clear relationship between relative growth and summer temperatures.From the study, we cannot conclude on any detrimental effects on growth from sulphur and acid deposition or from drought periods. A very recent study from the U.S., comprising 4800 plots and 24 tree species, confirms our results. However, we also show that the magnitude of N deposition effects on global forest C balance is currently a highly controversial matter, and comment on this debate. http://www.cef-cfr.ca/uploads/Colloque/Programme10_5.pdf

Abstract

The Lange Bramke catchment has been investigated as a monitored catchment for 60 years. However, its utilization history even dates back to medieval times, and is well documented in part. The intense interplay between ore mining, forestry, and water resources exploitation left remains such as scoriae piles and modified forest growth, e.g. due to local pollution at smelter locations. It is demonstrated that considering local land use history is important for a proper understanding and interpretation of modern monitoring data. A theoretical framework is proposed for the integration of the two data sources. This requires a joint approach combining two modelling paradigms, the functional one dominating in current ecosystem research, and an interactive one which best characterizes the human–environment relationship in historic times.

Abstract

The CO2 uptake of the terrestrial biosphere (Gross Primary Productivity, GPP) plays a key role in the global carbon balance. This carbon flux cannot be determined directly on a global scale. Yet, the remotely sensed Fraction of Absorbed Photosynthetically Active Radiation (FAPAR) is a valuable proxy for GPP. This study aims at characterizing global FAPAR dynamics on different temporal scales and extracting corresponding spatial structures. The time series were analyzed to uncover the presence and extent of trends, and to identify quasi-oscillatory patterns from intra- to interannual time scales....

Abstract

A catchment provides ecosystem data along with (relatively) simple, operationally defined boundaries. In addition runoff is an integrated measure of the hydrochemical ecosystem response, which can be represented by fluxes at the weir. Integration at the weir occurs first of all with respect to spatial scales. Almost all fluid output leaves a (tight) catchment at this point. Evaluation of the runoff dynamics (quantity and quality) is primarily concerned with temporal scales. The Lange Bramke catchment study with its four runoff series from forested catchments (spring and weir at Lange Bramke, weirs at Dicke Bramke and Steile Bramke) provides an exceptionally comprehensive data set. The following scales and processes can be considered, when interpreting temporal variations in runoff data: above the time scale of forest rotation (species composition, biomass accumulation, timber export, soil nutrient pools) at decadal time scales up to a full forest rotation of about 100-120 years (changes in forest growth rate, changes in deposition, climate change, insect outbreaks) at annual time scales (uptake, transpiration) at hourly to weekly time scales of hydrological events (precipitation, runoff, dilution effects of solvents).

Abstract

In high-latitude areas, landscapes with flat or moderate relief areas usually contain lakes and mires. The identification of flowpaths in such areas is a difficult issue. The increasing availability of high resolution topography from airborne Lidar measurements offers new opportunities for automatic or semi-automatic channel extraction from DEMs in small watersheds, substantially outperforming the hydrographic network in conventional digital maps....

Abstract

In Norway, it is planned to double the stationary use of bioenergy from all sources by up to 14 TWh before 2020, with much of this increase coming from forest resources, including residues like branches and tops (which are not much used today) being removed after tree harvest. This removal will reduce the supply of nutrients and organic matter to the forest soil, and may in the longer term increase the risk for future nutrient imbalance, reduced forest production, and changes in biodiversity and ground vegetation species composition. However, field experiments have found contrasting results (e.g. Johnson and Curtis 2001; Olsson et al. 1996). Soil effects of increased biomass removal will be closely related to soil organic matter (SOM) dynamics, litter quality, and turnover rates. The SOM pool is derived from a balance between above- and below-ground input of plant material and decomposition of both plants and SOM. Harvest intensity may affect the decomposition of existing SOM as well as the build-up of new SOM from litter and forest residues, by changing factors like soil temperature and moisture as well as amount and type of litter input. Changes in input of litter with different nutrient concentrations and decomposition patterns along with changes in SOM decomposition will affect the total storage of carbon, nitrogen and other vital nutrients in the soil. To quantify how different harvesting regimes lead to different C addition to soil, and to determine which factors have the greatest effect on decomposition of SOM under different environmental conditions, two Norway spruce forest systems will be investigated in the context of a research project starting in 2008/2009, one in eastern and one in western Norway, representing different climatic and landscape types. At each location, two treatment regimes will be tested: Conventional harvesting, with residues left on-site (CH) Aboveground whole-tree harvest, with branches, needles, and tops removed (WTH). Input of different forest residues will be quantified post harvest. Soil water at 30 cm soil depth will be analysed for nutrients and element fluxes will be estimated to provide information about nutrient leaching. Soil respiration will be measured, along with lab decomposition studies under different temperature and moisture regimes. Long term in situ decomposition studies will be carried out in the WTH plots using three different tree compartments (needles, coarse twigs, fine roots) decomposing in litter bags, in order to determine their limit value. The structure of the fungal community will be determined by soil core sampling and use of molecular techniques allowing qualitative and quantitative estimation. Understorey vegetation will be sampled to determine the biomass, and the frequency of all vascular plants, bryophytes and lichens will be estimated. After harvesting, replanting will be carried out. Seedling survival, causes of mortality and potential damage, growth, and needle nutrients will be monitored. Results from these studies will be used to identify key processes explaining trends observed in two series of ongoing long-term whole-tree thinning trials. We shall combine knowledge obtained using field experiments with results of modelling and data from the Norwegian Monitoring Programme for Forest Damage and National Forest Inventory. This will help us to predict and map the ecologically most suitable areas for increased harvesting of branches and tops on a regional scale based on current knowledge, and to identify uncertainties and additional knowledge needed to improve current predictions.

Abstract

Minirhizotrons, transparent acrylic tubes inserted in the soil, are well suited for long term, non destructive, in situ observations of fine roots. In minirhizotrons, the fine roots are regularly photographed and the root images are visually evaluated according to their status as living, dead or disappeared. This evaluation gives the background for further statistical treatment to estimate the fine root longevity. It is inherent in the minirhizotron technique that a large group of roots will be described as “disappeared” due to grazing, overgrowing by other roots, unclear images or other reasons. Because the fraction of disappeared roots is substantial in some cases, this has consequences for the interpretation of the longevity results. We processed three years of minirhizotron images from Norway spruce stands in southeast Norway (30 yr old) and northern Finland (60 yr old). Of all processed fine roots 32 and 23% was evaluated as disappeared in Norway and Finland, respectively. When roots labelled as disappeared were pooled together with dead ones, the fine root longevity estimates, using the Kaplan Meier method, decreased almost by a factor of two (401 and 433 days), as opposed to labeling them as censored observations (770 and 750 days for Norway and Finland, respectively). Here we demonstrate how the early decision making on the fine root status bears consequences on the resulting longevity estimates. The implications will be discussed

Abstract

Long-term monitoring meteorological, hydrological and hydrochemical data from small catchments are irreplaceable witnesses of past environmental conditions. This insight shaped the formation of the Long-Term Ecological Research Network (LTER) in the US, but European as well as German siblings are under preparation. Among the European forested monitoring sites, the Bramke catchments in the Harz mountains present a particularly well-documented case, with daily runoff measurements starting after World War II-related reparation cuts in 1948, and surface water chemistry being observed since the 1970ies. Originally powered by research on erosion, then by acid rain research and the then-prominent “forest decline”, a large set of hydrochemical variables (major ions) is available now with basically weekly resolution. Previously tightly connected to academic research at the University of Göttingen, routine measurements are by now performed by local forest authorities, ensuring forthcoming continuity even when public attention should shift away again from climate change research.

Abstract

Stream flow observations from a geographical region are known to often exhibit strong common behavior. In order to assess the temporal evolution of the complexity of stream flow time series from all over Europe, we employ the recently developed Linear Variance Decay dimension density (δLV D). It is estimated as the parameter of an exponential decay function fitted to the remaining variances of the eigenvalues of a covariance matrix. Scaling between zero and one, δLV D can be interpreted as a measure of the proportion of linear independent components in a multivariate record.....

To document

Abstract

In this study we demonstrate how airborne laser scanning (ALS) can be applied to map effective leaf area index (LAI(e)) in a spruce forest, after being calibrated with ground based measurements. In 2003 and 2005, ALS data and field estimates of LAI(e) were acquired in a Norway spruce forest in SE Norway. We used LI-COR's LAI-2000 (R) Plant canopy analyzer ("LAI-2000") and hemispherical images ("HI") for field based estimates of LAI(e). ALS penetration rate calculated from first echoes and from first and last echoes was strongly related to field estimates of LAI(e). We fitted regression models of LAI(e) against the log-transformed inverse of the ALS penetration rate, and in accordance with the Beer-Lambert law this produced a linear, no-intercept relationship. This was particularly the case for the LAI-2000, having R-2 values > 0.9. The strongest relationship was obtained by selecting ALS data from within a circle around each plot with a radius of 0.75 times the tree height. We found a slight difference in the relationship for the two years, which can be attributed to the differences in the ALS acquisition settings. The relationship was valid across four age classes of trees representing different stages of stand development, except in one case with newly regenerated stands which most likely was an artifact. Using LAI(e) based on HI data produced weaker relationships with the ALS data. This was the case even when we simulated LAI-2000 measurements based on the HI data. (C) 2009 Elsevier Inc. All rights reserved.

Abstract

Evaluation of climate change consequences and national carbon reporting such as under the Kyoto protocol require long-term monitoring of carbon fluxes. We report on an ongoing project aimed at a national-level assessment of the terrestrial carbon sequestration potential under present conditions and under various climate and land use change scenarios, in particular in terms of their temperature effect. We develop empirical models for national soil carbon stock assessment and evaluate process-based soil carbon models for prediction of future carbon dynamics.....

Abstract

General questions that arise while investigating hydrological extremes are whether these have distinct spatial and temporal variations and how these variations are linked to mean flow conditions. We analyze a large set of European stream flow series. Based on daily observations we derive annual series of stream flow deciles ranging from the minimum to the maximum, resulting in a set of eleven series for each station representing the year to year variability of the flow regimes....

Abstract

Life on earth depends on water and where running water occurs on earth, there is life. Nevertheless, existing modelling approaches in hydrology almost completely neglect the biological aspects of water flow. We claim that ignoring biological behaviour and interaction in catchment runoff modelling is too restrictive, and that computational theories can be used to formalise behaviour and interaction and model the biological impact on runoff. To demonstrate this, starting with a general classification of catchment behaviour, as documented in runoff data, we will use symbolic dynamics to quantify randomness and complexity in the time series. This approach shows that runoff records from very different catchments show common behaviour. This behaviour can be fitted to a one-parametric curve, stratified into three regions. In this manner, it becomes possible to represent and classify types of interactive behaviour that cannot be generated algorithmically. This suggests that physically based catchment models do not properly represent all types of interactive behaviour, and that signatures of biological interaction are present in runoff data.

Abstract

Over the last decades the forestry sciences have been opened for new topics and methods. In addition to traditional forestry topics they have participated in environmental and ecosystem research. So far this type of research has been perceived as “applied”. From the modelling perspective there has been a misunderstanding among the participating disciplines of the character of knowledge being applied. Here we introduce two types of models of forest utilization and discuss their possibilities and limits for forestry sciences. The first perspective of forests is the one dominating in modelling today and in forestry sciences. It has been adopted from physics. The second perspective of forests has implicitly been adopted in the past for pragmatic reasons.

Abstract

Many time series analysis methods depend on equally spaced observations with no data point missing. If this condition is met, powerful techniques are available that identify temporal structures such as trends or periodic phenomena or nonlinear dynamics. Unfortunately, most of observations of natural systems, in particular over longer periods of time such as decades, are prone to sampling errors leading to missing points in the observations. Singular System Analysis (SSA) is a powerful tool to extract the dynamics contained in time series at arbitrary temporal scales...

Abstract

Runoff time series are known to contain long term structures on interannual to decadal time scales. Investigating spatial patterns of long term structures is a way to elucidate the relationship between external forcings and watershed properties. This would be a valuable contribution to an improved water resources management. Singular System Analysis (SSA) is a powerful technique to identify and extract significant long term components from time series. However, many observations from natural systems are prone to missing data that hamper many analysis techniques, including the SSA in its original formulation...

Abstract

Insect-induced damages in forests are a major concern for timber production, landscape conservation and ecosystem research. Early detection methods based on remote sensing data can document the severity and spatial extent of ongoing attacks and might aid in designing mitigation measures or even prevention where necessary. In southeastern Norway, a large-scale insect defoliation of pine trees is ongoing. The larvae of the Pine sawfly Neodiprion sertifer reate it with its mass attacks during their feeding on needles in June and July. In the winter before the attack, egg galleries are evident in the needles. This provides a test case for early detection methods and remote sensing techniques for monitoring forest health....

Abstract

We investigated whether the stand age affects the life span of tree and understory fine roots (<1mm) in three Norway spruce (Picea abies) stands: 30, 60 and 120-yr-old. In each stand 9 minirhizotrons were installed and images were collected once in a month throughout the growing season during the three years. Norway spruce fine roots in the 30-yr old stand had a life span 401 ± 27 and 341 ± 68 days, and understory 409 ± 162 and 349 ± 142 days, estimated by using the Kaplan Meier survival analysis (KM) and Weibull distribution, respectively...

Abstract

In European forests, standings stocks are currently increasing and are higher than ever during the last decades. This is due to a multitude of reasons; human impacts such as reduced logging or the abandonment of agricultural land are clearly among them. However, data from intensive monitoring plots reveal an increased growth even in the absence of direct human intervention. For this study, we used a set of 363 such plots from 16 European countries, which are a subset of the ICP-Forests Level II plots, and are typically rectangular areas with a size of 0.25 ha. We investigated the influence of environmental factors on forest growth. In particular, the role of nitrogen, sulphur and acid deposition, temperature, precipitation and drought was elucidated. The study focussed on the tree species Norway spruce, Scots pine, common beech and European as well as sessile oak. We used existing information on site productivity, stand age and stand density to estimate expected growth. Relative tree growth was then calculated as the ratio between actual growth, obtained within a five years observation period, and expected growth. The site productivity incorporates past environmental conditions and was either computed from site index curves, where we distinguished Northern, Central and Southern Europe variants, or was taken from expert estimates. The models explained between 18% and 39% of the variance. Site productivity and stand age were positively and negatively related to actual growth, respectively. The results indicated consistently a fertilizing effect from nitrogen deposition, with roughly one percent increase in site productivity per kg of nitrogen deposition per ha and year, most pronounced for plots having soil C/N ratios above 25. We also found a positive albeit less clear relationship between relative growth and summer temperatures. Other influences were uncertain. In particular, we cannot conclude on detrimental effects on growth from sulphur and acid deposition or from drought periods.

Abstract

River floods may cause considerable damage. Water management strategies intend tomoderate or mitigate the severe effects of extreme discharge events. In this context,techniques for the detection and attribution of changes is of crucial importance. Extremeflood events seem to occur more frequently in recent decades in central Europe.It is anticipated that climate change and weather regime shifts may contribute to this...

Abstract

Characterizing ecosystem-atmosphere interactions in terms of carbon and water exchange on different time scales is considered a major challenge in terrestrial biogeochemical cycle research. The respective time series currently comprise an observation period of up to one decade. In this study, we explored whether the observation period is already sufficient to detect cross-relationships between the variables beyond the annual cycle, as they are expected from comparable studies in climatology. We investigated the potential of Singular System Analysis (SSA) to extract arbitrary kinds of oscillatory patterns. The method is completely data adaptive and performs an effective signal to noise separation. We found that most observations (Net Ecosystem Exchange, NEE, Gross Primary Productivity, GPP, Ecosystem Respiration, Reco, Vapor Pressure Deficit, VPD, Latent Heat, LE, Sensible Heat, H, Wind Speed, u, and Precipitation, P) were influenced significantly by low-frequency components (interannual variability). Furthermore, we extracted a set of nontrivial relationships and found clear seasonal hysteresis effects except for the interrelation of NEE with Global Radiation (Rg). SSA provides a new tool for the investigation of these phenomena explicitly on different time scales. Furthermore, we showed that SSA has great potential for eddy covariance data processing, since it can be applied as a novel gap filling approach relying on the temporal correlation structure of the time series structure only.

Abstract

Ecological studies are often confronted with short and fragmented or unevenly sampled time series. Examples are, e.g., time series of biogeochemical fluxes measured on a variety of scales. Characterizing the observed time series patterns, particularly the correlation structure is crucial for an integrated ecosystem assessment or possibly for improved processes understanding.

Abstract

In the context of an ongoing project on REMote sensing of FORest health (REM-FOR), we analyze airborne high-resolution hyperspectral images of a pine-dominated region in southeast Norway heavily attacked by the Pine sawfly Neodiprion sertifer, leading to severe defoliation. Leaf Area Index (LAI) is used as a proxy of the crown density, and comparing LAI maps from before and after the attack lead to indicators for damage extent. We discuss the application of the Forest Reflectance Model (FRT) of Kuusk and Nilson, which was designed for the application to (managed) Northern European Forests, to model the spectral response from the canopy. It is based on conventional forest inventory data, species-dependent parametrized crown shapes, canopy LAI, needle clumping index, and needle optical properties. Here, however, we run the model in an inverse mode, by iteratively minimizing the discrepancy between measured and simulated reflectances, and predicting the LAI, keeping known parameters of the model fixed. The LAI values are then compared to those obtained with either ground-based Licor LAI2000 measurements, or with airborne laser-scanning. Some preliminary results of this modelling concept for the case study are discussed.

Abstract

The net ecosystem productivity (NEP) of a sequence of threemonoaged Norway spruce stands located in southeast Norway is modelled using the biogeochemical model Biome-BGC. For calibration, we use estimated biomass stocks at the plot level and Leaf Area Index measurements. The model is run for 30 years of historical temperature measurements as well as for a regional climate scenario. It is shown that under current conditions, NEP develops from negative values for a young stand (30 years) to clearly positive for a middleaged (60 years) to slightly negative again for a very old and decaying stand (120 years). However, the old stand benefits substantially from the predicted increased temperatures in the climate scenario, rendering NEP positive again. For the 30 and 60 years stands, almost no change is predicted from Biome-BGC.

Abstract

The REMFOR project evaluates remote sensing data and methods for monitoring forest health using variation in leaf area index (LAI) as a primary measure of defoliation. A large-scale pine sawfly outbreak in Norway serves as a test case. An LAI map of the study area was derived from airborne LIDAR measurements before and after the insect attack to serve as ground truth for satellite image analysis. The method predicts LAI from laser penetration rates through the canopy layer in accordance with the Beer- Lambert law calibrated with point measurements of LAI with LICOR LAI-2000. Comparing two cloud-free SPOT scenes from September 2004 and September 2005 shows obvious visual patterns of defoliation in pine forests from the 2005 outbreak. Preliminary analysis shows that the insect defoliation caused an increase in middle-infrared (SPOT band4) reflectance and a decrease in SPOT NDVI, and both these responses may be used as a reasonable predictor of LAI loss as derived from laser scanning. MODIS NDVI data were gathered for the area over the period 2000-2006, and the Timesat algorithm is used to smooth the seasonal variation. The insect attack is evident from the smoothed NDVI data both as a reduction in the summer mean value, and as an alteration of the seasonal profile during the larvae feeding period in June and July. REMFOR also encompasses a range of other remote sensing data types, including GLAS LIDAR, SAR and hyperspectral data from both airborne and satellite platforms (e.g. Hyspex and Hyperion). Landsat TM is used to generate a tree species map.

Abstract

Model simulations show that an increased frequency in storms and drought periods may result in more frequent and shorter outbreaks of bark beetles. Warmer summers can result in two bark beetle generations per summer instead of one, giving bark beetles the opportunity to attack forests twice in a single year.

Abstract

Toxic effects of aluminium (Al) on Picea abies (L.) Karst. (Norway spruce) trees are well documented in laboratory-scale experiments, but field-based evidence is scarce. This paper presents results on fine root growth and chemistry from a field manipulation experiment in a P. abies stand that was 45 years old when the experiment started in 1996. Different amounts of dissolved aluminium were added as AlCl3 by means of periodic irrigation during the growing season in the period 19972002. Potentially toxic concentrations of Al in the soil solution were obtained. Fine roots were studied from direct cores (1996) and sequential root ingrowth cores (1999, 2001, 2002) in the mineral soil (040 cm). We tested two hypotheses: (1) elevated concentration of Al in the root zone leads to significant changes in root biomass, partitioning into fine, coarse, living or dead fractions, and distribution with depth; (2) elevated Al concentration leads to a noticeable uptake of Al and reduced uptake of Ca and Mg; this results in Ca and Mg depletion in roots. Hypothesis 1 was only marginally supported, as just a few significant treatment effects on biomass were found. Hypothesis 2 was supported in part; Al addition led to increased root concentrations of Al in 1999 and 2002 and reduced Mg/Al in 1999. Comparison of roots from subsequent root samplings showed a decrease in Al and S over time. The results illustrated that 7 years of elevated Altot concentrations in the soil solution up to 200 M are not likely to affect root growth. We also discuss possible improvements of the experimental approach.

Abstract

Ecosystems commonly fall under the rubric of complex systems (West and Brown 2004). Nevertheless, in the practical management of certain ecosystems, we encounter simple heuristic rules of human interference that are often derived from cultural traditions rather than from scientific study. The increased technical power of computer-based simulation tools and their increased mathematical formalization may either remove former technical limits (e.g., of prediction) or, in contrast, reveal the fundamental character of some of these limits. Here, we shall argue that both cases occur, and that the main effect of simulation technology is to bring the distinction between these cases into scientific awareness.

Abstract

The minirhizotron technique provides the opportunity to perform in situ measurements of fine root dynamics and obtain accurate estimates of fine root production and turnover. The objective of the present work was to determine the fine root longevity and mycorrhization in a Norway spruce chronosequence. The study was carried out on four stands of planted Norway spruce (Picea abies), approximately 10, 30, 60 and 120 years old, during 2001 and 2002. The stands were located at Nordmoen, a plain of sandy deposits in southeast Norway (60o15 N, 11o06 E). For the root turnover study, altogether 60 minirhizotrones were installed and images were processed.Individual fine roots were identified, their mycorrhization assessed, appearance and possible disappearance dated, and growth in length measured. The data set was subjected to a survival analysis, using a Kaplan-Meier product-limit approach. The minirhizotron samples were stratified according to stand age class, and Coxs F-test was used to analyze differences in survival estimates. The analysis may also be extended to consider other covariates such as tree species (spruce, pine or birch), understory vegetation, or soil depth. Typical survival function estimates will be presented, and the influence of stand age on the mycorrhization and the dynamics of the fine roots will be discussed.

Abstract

Most phenomena in ecosystem research are assessed via repeated measurements of environmental variables. The dynamics of these time series is investigated with a variety of statistical techniques; in this article, we focus on modern nonlinear methods. They enable separation of short- and long-term components, show all types of trends and quantify the information contained and the complexity of the data sets.

Abstract

Considerable knowledge exists about the effect of aluminium (Al) on root vitality, but whether elevated levels of Al affect soil microorganisms is largely unknown. We thus compared soils from Al-treated and control plots of a field experiment with respect to microbial and chemical parameters, as well as root growth and vitality.The field experiment was established in a 50-year-old Norway spruce (Picea abies L.) stand where no Al or low concentrations of Al had been added every 710 days during the growth season for 7 years. Analysis of soil solutions collected using zero tension lysimeters and porous suction cups showed that Al treatment lead to increased concentrations of Al, Ca and Mg and lower pH and (Ca Mg K)/(Al) molar ratio. Corresponding soil analyses showed that soil pH remained unaffected (pH 3.8), that exchangeable Al increased, while exchangeable Ca and Mg decreased due to the Al treatment.Root in-growth into cores placed in the upper 20 cm of the soil during three growth seasons was not affected by Al additions, neither was nutrient concentration or mortality of these roots. The biomass of some taxonomic groups of soil microorganisms, analyzed using specific membrane components (phospholipid fatty acids; PLFAs), was clearly affected by the imposed Al treatment, both in the organic soil horizon and in the underlying mineral soil.Microbial community structure in both horizons was also clearly modified by the Al treatment. Shifts in PLFA trans/cis ratios indicative of short term physiological stress were not observed. Yet, aluminium stress was indicated both by changes in community structure and in ratios of single PLFAs for treated/untreated plots. Thus, soil microorganisms were more sensitive indicators of subtle chemical changes in soil than chemical composition and vitality of roots.

Abstract

We investigate ecosystem dynamics by analyzing time series of measured variables. The information content and the complexity of these data are quantifed by methods from information theory.When applied to runoff (stream discharge) from catchments, the information/complexity relation reveals a simple non-trivial property for a large ensemble (more than 1800) of time series. This behaviour is so far not understood in hydrology.Using a multi-agent network receiving input resembling rainfall and producing output, we are able to reproduce the observed behaviour for the first time. The reconstruction is based on the identification and subsequent replacement of general patterns in the input. We thus consider runoff dynamics as the expression of an interactive learning problem of agents in an ecosystem.

Abstract

We investigate a data set of 160 river runoff time series at daily resolution from catchments in Southern Germany. Our aim is to seek spatial patterns for best parametrization of extreme value distributions to these data sets on one hand, and to analyze temporal instationarities of parameter estimates and extreme value attributes on the other. Conventional extreme value statistics and the calculation of return periods implicitly assume that the most extreme events are statistically independent. We demonstrate that this assumption is invalid, and that correlations, temporal as well as spatial, of arbitrary extent prevail instead. An important consequence is that the concept of return periods is obsolete. In order to find explanatory variables for the observed patterns, features of the waiting time distribution at a given relative threshold are correlated to catchment properties, such as size, mean runoff volume, elevation, and others. Finally, the effect of varying temporal resolution on the duration periods is exhibited. http://www.cosis.net/abstracts/EGU05/03192/EGU05-J-03192.pdf

Abstract

Instationarities in runoff time series are ubiquitous. However, simple trend analyses are often obscured by the presence of long-term correlations, and some instationarities are not simply changes in the mean or periodicities. Thus, wherever feasible, instationarities should be based on the full frequency distribution, or the cumulative distribution function (cdf), of the series. In this paper, we investigate the time-dependence of the empirical cdfs of 97 runoff datasets from the upper Danube basin applying a new pairwise test statistic, KSSUM, based on integrated differences of the cdfs. This is an improvement to the Kolmogorov-Smirnov (KS) test and was applied on different time scales, i.e. windows of varying size. If desired, the influence of drifts in the mean as well as heteroscedasticity can be excluded via z-transformations. The resulting time series of the KSSUM variable, either within a runoff series for different windows, or across series for the same period, is then subjected to the detection of spatiotemporal patterns with different methods. For most of the time series the underlying distributions move towards higher values in the long run. We also observed a periodic drift in the mean across all analysed gauges. It is furthermore possible to separate exceedingly variable runoff series from those with intermediate or small changes in value distribution on a regional basis, and thus to separate overall trends from local deviations at individual gauges. It is demonstrated that KSSUM is a sensitive method to investigate instationarities in sets of time series based on pairwise comparisons. An extension to a proper multivariate comparison is a possible further development. http://www.cosis.net/abstracts/EGU05/04198/EGU05-J-04198.pdf

Abstract

Root and needle litter are the most important sources of organic carbon in forest soils. Their decomposition is thus important for the long-term storage of C in, and release of CO2 from, the soil. Different components in the organic matter will decompose with different speeds. NIRS (Near InfraRed Spectroscopy) is a relatively simple and promising way of analysing the composition of organic matter, but its use in forest soil and litter studies has been limited up to now. We will present preliminary results from litter decomposition studies in two forest ecosystems: Picea abies stands (30 and 120 years old) from Nordmoen, Norway, and uneven-aged P. abies stands with a mean age of 90 years and under different N treatments at Gårdsjön, Sweden. ags with litter collected from the stands have been buried in the soil for different time periods and have been analysed using a CHN-analyzer and NIRS. Two aspects will be discussed: a) model calibration and validation for C and N concentrations, and b) assessment of decomposability using NIRS.

Abstract

Conventional extreme value statistics and the calculation of return periods implicitly assume stationarity of distributions and statistical independence at least asymptotically (most extreme events).We demonstrate, using a collection of river runoff time series from Southern Germany, that these assumptions are invalid, and that temporal as well as spatial correlations prevail instead: temporal differences of distributions are nearly synchronized within a region, and there are systematic trends of percentiles especially at low flow conditions within the 20th century.As a consequence, the estimated return periods of a given threshold flow are fluctuating, in some cases even in a dramatic fashion. On the other hand, a general trend towards an increase in flood frequencies cannot be stated on basis of our investigations, in accordance with other recent findings (Mudelsee et al. 2003), but contrary to general expectations drawn from climate change studies.

Abstract

Sampling the catchment outlet generally is assumed to be a convenient way to infer information about a variety of biogeochemical processes at the catchment scale as it provides a spatial and temporal integral of the predominating catchment output fluxes for a number of chemical compounds of interest.Moreover, the short-term dynamics and long-term trends of the hydrograph and of solute concentrations in the catchment runoff can provide information about the predominating processes at the catchment scale and can be used to refine conceptual and mathematical models.Additional measurements inside the catchment, e.g., of soil solution, groundwater, and stream water at different sites, are used to relate the findings to within-catchment processes and thus to further constrain hypotheses and models.

Abstract

Using Singular System Analysis (SSA), we extract a collection of significant long-term components (with dominant periods of at least 3 years) for a large number of river runoff records.At first glance, these long-term modes are a distinct feature of this variable, not contained in precipitation and temperature, and not easily correlated to commonly known long-term indices (NAO, SOI, NHT, SUN, etc.). However, low-pass filtered versions of these time series exhibit strikingly similar behavior, like common maxima, within a region (such as Southern Germany), pointing to a common origin.Although not an unequivocal example for synchronization, we quantify the degree of synchronization as a function of the regional extent of the data and propose a mechanism, stochastic resonance, discussed in climate dynamics, which is able to produce this collective behavior despite the lack of deterministic drivers. We also comment on air pressure-induced teleconnections between the different large scale oscillations in the climate system.

Abstract

Ecosystems as objects of natural sciences are often difficult to understand, as an object of traditional management they are sometimes easy to utilize. Computer-based modeling offers new tools to study this apparent paradox.We propose an interactive framework from which the traditional approach based on dynamic system theory can be challenged for living systems: Models derived on the basis of the state concept have not (yet?) allowed predictions that derive novel management competence relevant for the altered boundary conditions of ecosystems. Here a concept of interaction as currently used in information sciences serves as starting point for deriving models more appropriate for ecosystems.An application and test of this concept consists in a search for signatures of interaction in environmental and ecological time series. Confronted with the notorious lack of detailed process understanding, it is plausible to rely on time series analysis techniques. The intricate nature of typical multivariate data sets from ecosystems immediately suggests a preference for nonlinear techniques, and among them temporally local methods, able to detect even subtle changes in the underlying dynamics.We shortly introduce a couple of these methods, which have been demonstrated to be appropriate for time series exceeding minimal length requirements. This is exemplified by recurrence quantification analysis. In addition we present methods to quantify the memory content (Hurst analysis) and complexity of data sets (defined in an informationtheoretic context).Time series analysis of extended environmental and ecological data sets can give detailed structural insights, monitors subtle changes undetectable otherwise, forms the basis for further inferences and provides rigorous model testing on all scales. The success of dynamic system theory when applied to non-living environmental data is strikingly contrasted by the difficulties of the same method when dealing with ecological data We conjecture that this difference reflects the extent to which interaction has been disregarded for ecosystems.

Abstract

Extensive monitoring of forest health in Europe has been carried out for two decades, based mainly on defoliation and discolouration. Together these two variables reflect chlorophyll amounts in the tree crown, i.e. as an indicator of foliar mass, and chlorophyll concentration in the foliage, respectively.In a current project we try to apply remote sensing techniques to estimate canopy chlorophyll mass, being a suitable forest health variable. So far, we limit this to Norway spruce only. LIDAR data here play an important role, together with optical and spectral data, either from survey flights or from satellites. We intend to model relationships between foliar mass and LIDAR data for sample trees, and then scale up this to foliar mass estimates for the entire LIDAR area.Similarly, we try to scale up chlorophyll concentrations in sample trees, by modelling a relationship between sample tree chlorophyll and hyper-spectral data. The estimates of foliar mass and chlorophyll concentrations are then aggregated to every 10x10 m pixel of a SPOT satellite scene which is also covered by airborne data, providing an up-scaled ground truth. If we are successful with this, it might be a starting point for developing a new nationwide forest health monitoring system in Norway.

Abstract

The process of model building in the environmental sciences and when dealing with ecosystems is discussed. Two types of modeling approaches need to be distinguished: An algorithmic one, which has been used traditionally in physics, meteorology, and other branches where biological degrees of freedom are either absent or neglectable; and an interactive one, which is a new framework in computer science and seems to be most suitable in cases where organisms (including humans) as agents in ecosystems are to be taken into account. The first modeling approach is exemplified by state models in dynamic systems theory and expresses the correspondence imposed by Natural Law between inferential entailment in a formal system and causal entailment in natural systems. Modeling is to be separated from simulation. Simulation is a less restrictive type of modeling in which the description of non-interactive behaviour is the purpose and no constraints on the correspondence to internal states are imposed. The second (new) modeling approach is exemplified by interactive simulation models. It is able to express the correspondence in behaviour imposed by engineering standards (or cultural norms in general) between documentation, training and application in interactive choice situations such as games or ecosystem management. It generalises the notion of simulation for interactive problems. In an idealised situation the strictest correspondence between behaviour in a natural and a virtual system is expressed as bisimulation. The principles for model building are shortly demonstrated with examples.

Abstract

The forest stand growth simulator TRAGIC (tree response to acidification in groundwater in C) which has been developed to serve as a decision support system and a visualisation tool for scientists and forestry practitioners is introduced. TRAGIC places an emphasis upon visualisation techniques while at the same time providing detailed information on tree physiology and related parameters. The model is calibrated numerically to growth history data from two different European sites.Next, due to the importance of the visual component of the model, its ability to reproduce forest stand spatial structure is investigated, using an application of the theory of marked point processes. This analysis is applied to different experimental data sets for stands of different age, revealing information on planting schemes and the extent of significant spatial correlations.The spatial structure of the two model calibrations is then explored with the same methods. The point process analysis turns out to be a powerful diagnostic for model quality assessments, since spatial distribution is an indirect result of competition between trees for light.

Abstract

The notion of an ecological damage has so far neither been given a proper theoretical nor a pragmatic or operational foundation. Yet one of the most widespread motivations for the scientific study of ecosystems is a “protectional” one by which an improved scientific understanding is sought in order to be able to prevent future ecological damages. We review the possibilities of valuating changes in the environment, in health or in ecosystems as a damage. The conceptual separation of potential from actual behaviour/structure is a prerequisite to any of them. The critical point here is the formal and empirical basis for the knowledge about these potentials. We contrast the dynamic systems theory approach derived in physics with an interactive computing approach recently developed in computer science. The former requires to distinguish facts and values and leads to notorious difficulties when applied at the ecosystem level. The latter and novel approach opens the possibility for a consistent definition of a damage at the ecosystem level whenever a tradition of (sustainable) utilization of such systems is available. The documentation, actualisation and dissemination of the tacit (expert) knowledge can be improved by the use of interactive simulations in which a virtual standard can defined by the respective experts themselves.

Abstract

The rationale for stand growth modelling is often either grounded in a search for improved scientific understanding or in support for management decisions. The ultimate goal under the first task is seen in mechanistic models, i.e. models that represent the stand structure realistically and predict future growth as a function of the current status of the stand. Such mechanistic models tend to be over-parameterized with respect to the data actually available for a given stand. Calibration of these models may lead to non-unique representations and unreliable predictions. Empirical models, the second major line of growth modelling, typically match available data sets as well as do process-based models. They have less degrees of freedom, hence mitigate the problem of non-unique calibration results, but they employ often parameters without physiological or physical meaning. That is why empirical models cannot be extrapolated beyond the existing conditions of observations. Here we argue that this widespread dilemma can be overcome by using interactive models as an alternative approach to mechanistic (algorithmic) models. Interactive models can be used at two levels: a) the interactions among trees of a species or ecosystem and b) the interactions between forest management and a stand structure, e.g. in thinning trials. In such a model data from a range of sources (scientific, administrative, empirical) can be incorporated into consistent growth reconstructions. Interactive selection among such growth reconstructions may be theoretically more powerful than algorithmic automatic selection. We suggest a modelling approach in which this theoretical conjecture can be put to a practical test. To this end growth models need to be equipped with interactive visualization interfaces in order to be utilized as input devices for silvicultural expertise. Interactive models will not affect the difficulties of predicting forest growth, but may be at their best in documenting and disseminating silvicultural competence in forestry.

Abstract

Reprints available in my office

Abstract

Living organisms in ecosystems are conceptualized as autonomous agents with a spectrum for their behavior. Ecosystems are described here as interacting multi-agent systems. Implementing such a system is a challenge for current hardware and software technology both technically and conceptually, in particular if one of the agents is human, either virtually within the system or as external participant and user (real human).Interfering with and manipulating the system occurs at arbitrary times during simulation, with a collection of choices to do that, rendering the details of the particular simulation fundamentally unpredictable.As a result, we have an open interactive system with tight feedback loops, for which new computer models (beyond the Universal Turing Machine) are required. We discuss some of the theoretical concepts for the appropriate software technology and shortly present one example of such a system, a forest simulator used by forest administrators.

Abstract

High resolution digital elevation maps (DEMs) offer the investigation of multifractal properties of the spatial characteristics of river basins like the width function, and the determination of the relation between average slope and basin area.There have been a number of universality claims in this respect; the range of the scaling exponent for the slope-area relation seems to be narrow, and the multifractal spectrum of the width function is characterized by a single site-specific Lipschitz-Hlder exponent alpha, the spectrum having an envelope given by that of Peanos basin.Comparing 17 river basins covering two orders of magnitude in basin area, our findings do not confirm this universal character. In particular, the Lipschitz-Hlder exponent crucially depends on the resolution of the width function extraction; we show that it is easy to produce almost identical spectra for completely different basins when varying the resolution.The problem of interior points is also encountered. We adopt Venezianos modified calculation of f(alpha) in this case. The slope-area exponent covers a wide range of values which also include the pure random case. We thus question the usability of these measures as a classification tool for river basins. http://www.cosis.net/abstracts/EAE03/05246/EAE03-J-05246.pdf

Abstract

For the intepretation of multifractal properties of experimental time series, two prominent procedures used are the double trace method (DTM) and the universal multifractal (UM) approach. We calculated multifractal spectra for a collection of long-term precipitation, air temperature and river discharge records, covering a wide range of spatial scales.Considering K(q) in this framework leads to an effective classification of dynamical behavior. Comparison of the DTM and UM methodologies, however, reveals substantial differences which make them difficult to reconcile. This is in particular true for the discharge case.The scaling exponent is generally larger in magnitude for the DTM and in some cases even extends into the non-analytical regime. Part of previous work thus could not be confirmed. Whether the description of river flow as multifractal process is feasible remains an open question. http://www.cosis.net/abstracts/EAE03/05092/EAE03-J-05092.pdf

Abstract

An individual-based agent model is presented which resembles aspects of natural evolution in ecosystems under selective pressure due to limited resources. The environmental conditions are determined by spatial and temporal variability of resource abundances.The agents have to choose between three different types of resources; the one consumed most during lifetime solely counts for the fitness of the individual agent. Simulation runs show that populations specialized in different resource types are mutually influencing each other under temporal variation of a single resource type.Mobility of agents in a locally heterogenous world enables recolonization after a population has starved to death. Wavelet analysis of the population time series reveals that some observed population dynamics show phenomena such as localized periodicities which cannot be explained by linear dependencies on the resource input dynamics.

Abstract

Modern information technology allows the investigation of the characteristic properties of living systems from a new perspective. Which of the ecosystem features are necessary conditions resulting from their constraints, which are accidental, constituting contingent facts of their respective histories?As long as we know of a single phylogenetic tree in nature, the difference is hard to tell, rendering the reconstruction and realisation of artificial ecologies a major challenge. It has been taken up by the high technology of the time since decades; since two decades, IT is leading in this respect.Are there life forms that can be created in contemporary computers, and which ones? Successes and failures of a number of virtualizations are forming de facto constraints for theoretical ecosystem research. Artificial Life (AL) research appears to be not just another attempt towards realistic models for ecological systems, but undermines the basic assumptions of most of conventional modeling in this area: in AL, behavior is in general irreducible to internal mechanisms; behavior results rather from interactive and intentional usage of the simulation.We try to elucidate and demonstrate the crucial role of interaction in these simulations, drawing from current developments in theoretical computer science as well as a number of examples. We propose a new classification of ecosystem models according to its degree of interactivity.