Above-Ground Biomass (AGB) assessment using remote sensing has been an active area of research since the 1970s. However, improvements in the reported accuracy of wide scale studies remain relatively small. Therefore, there is a need to improve error analysis to answer the question: Why is AGB assessment accuracy still under doubt? This project aimed to develop and implement a systematic quantitative methodology to analyse the uncertainty of remotely sensed AGB, including all perceptible error types and reducing the associated costs and computational effort required in comparison to conventional methods. An accuracy prediction tool was designed based on previous study inputs and their outcome accuracy. The methodology used included training a neural network tool to emulate human decision making for the optimal trade-off between cost and accuracy for forest biomass surveys. The training samples were based on outputs from a number of previous biomass surveys, including 64 optical data based studies, 62 Lidar data based studies, 100 Radar data based studies, and 50 combined data studies. The tool showed promising convergent results of medium production ability. However, it might take many years until enough studies will be published to provide sufficient samples for accurate predictions. To provide field data for the next steps, 38 plots within six sites were scanned with a Leica ScanStation P20 terrestrial laser scanner. The Terrestrial Laser Scanning (TLS) data analysis used existing techniques such as 3D voxels and applied allometric equations, alongside exploring new features such as non-plane voxel layers, parent-child relationships between layers and skeletonising tree branches to speed up the overall processing time. The results were two maps for each plot, a tree trunk map and branch map. An error analysis tool was designed to work on three stages. Stage 1 uses a Taylor method to propagate errors from remote sensing data for the products that were used as direct inputs to the biomass assessment process. Stage 2 applies a Monte Carlo method to propagate errors from the direct remote sensing and field inputs to the mathematical model. Stage 3 includes generating an error estimation model that is trained based on the error behaviour of the training samples. The tool was applied to four biomass assessment scenarios, and the results show that the relative error of AGB represented by the RMSE of the model fitting was high (20-35% of the AGB) in spite of the relatively high correlation coefficients. About 65% of the RMSE is due to the remote sensing and field data errors, with the remaining 35% due to the ill-defined relationship between the remote sensing data and AGB. The error component that has the largest influence was the remote sensing error (50-60% of the propagated error), with both the spatial and spectral error components having a clear influence on the total error. The influence of field data errors was close to the remote sensing data errors (40-50% of the propagated error) and its spatial and non-spatial Overall, the study successfully traced the errors and applied certainty-scenarios using the software tool designed for this purpose. The applied novel approach allowed for a relatively fast solution when mapping errors outside the fieldwork areas.

Metodologias que sejam padronizadas e consolidadas para quantificar carbono em florestas tropicais vem sendo discutidas em convenções climáticas. Este trabalho contribui para a estimativa de biomassa e carbono estocado na parte aérea de áreas de restauração da Mata Atlântica em torno de reservatórios localizados no Pontal do Paranapanema utilizando tecnologia LiDAR (Light Detection and Ranging). Procurou-se explorar o acúmulo e o estoque de carbono em três florestas em diferentes condições de sucessão: duas florestas caracterizdas como madura e a floresta restaurada. Na primeira etapa do trabalho, houve a escolha das equações alométricas encontradas em literatura para determinar a quantidade de carbono em cada uma das áreas utilizando variáveis medidas em campo. Assim, foi utilizada a equação de Ferez et al. (2015), ajustada em floresta restaurada da Mata Atlântica, para quantificar o carbono da área de estudo. Na segunda etapa do trabalho, procurou-se estimar a taxa de incremento do estoque de carbono da parte aérea, utilizando as variáveis medidas em campo obtidas em duas campanhas, nos anos de 2015 e 2016. O corredor restaurado apresentou média de 7,1 Mg.C.ha-1, a floresta madura da ESEC apresentou 39,9 Mg.C.ha-1, e a floresta madura do Morro do Diabo apresentou 45,2 Mg.C.ha-1 para o ano de 2016. A fixação anual encontrada na variação do estoque de 2015 para 2016 foi de 1,2 Mg.ha-1.ano-1 para a floresta restaurada, 1,6 Mg.ha-1.ano-1 para a floresta madura da ESEC e 2.5 Mg.ha-1.ano-1 para a floresta madura do MD. A terceira etapa do trabalho traz a modelagem realizada com dados LiDAR e dados do inventário convencional. Após utilizados métodos estatísticos para seleção de modelo, coeficiente de determinação ajustado, critério de Akaike e erro padrão da estimativa, o modelo escolhido utiliza as métricas percentil 90 e porcentagem de retornos acima de 50cm do solo para estimar carbono acima do solo, obtendo como coeficiente de determinação 0,78. A extrapolação para a área total pelo método do inventário convencional e pelo método da modelagem LiDAR apresentaram diferenças, demonstrando a utilização do LiDAR para reconhecer e retirar informações de áreas não amostradas. O quarto capítulo dessa dissertação traz a variação de carbono estimado pela modelagem com métricas LiDAR para os anos de 2015 e 2016, além do mapa do estocagem evidenciando os locais de maior sucesso e os de menor. A tecnologia LiDAR se mostrou eficiente em captar a variação ocorrida no intervalo de um ano e na quantificação de carbono. / The establishment of methodologies used to quantify carbon in tropical forests are one of the main topics on climate conventions. This project contributes to the estimation of biomass and aboveground carbon stored around reservoirs located in Pontal do Paranapanema, São Paulo - Brazil, using Light Detection and Ranging (LiDAR) technology. The objective was to explore the accumulation of aboveground carbon stored in three different succession conditions: mature, secondary and restored forest. The first chapter of this thesis shows the common allometric equations found in literature used to determine the amount of carbon in each area using field variables. However, the chosen allometric equation was developed by Ferez et al. (2015), because it was adjusted in similar areas than the area of this work. Thus, the restored corridor presented a mean of 7.1 Mg.C.ha-1, the secondary forest 39.9 Mg.C.ha-1 and the mature forest 45.2 Mg.C.ha-1 for the year of 2016. The annual fixation found was about 1.2 Mg.ha-1 for the restored forest, 1.6 Mg.ha-1 for the secondary forest and 2.5 Mg.ha-1 for the mature one. The second article brings the modeling performed with LiDAR data and traditional field inventory data. After using statistical methods for model selection, the chosen model uses two metrics: percentile 90 and percentage of returns above 50 cm of height to estimate aboveground carbon The extrapolation to the total area by traditional inventory and LiDAR modeling showed differences, demonstrating the efficiency of LiDAR to recognize information from non-sampled areas. The last chapter of the thesis brings the variation of carbon using LiDAR data through the years of 2015 and 2016. LiDAR data showed to be useful for measuring aboveground carbon and to detect the increment of the carbon stock over a year.

The dryland ecosystems of the US Desert Southwest (SW) are dependent on soil moisture for aboveground productivity; the generation of soil moisture in the SW is dependent on both soil physical properties and climate forcing. This study is one of the first regional point-scale analyses that explores the role of soil physical properties in modulating aboveground vegetation dynamics in response to climate forcing in the SW. Soil texture accounted for significant differences in average aboveground primary productivity across the SW. However, soil texture could not account for differences in inter-annual aboveground productivity variation across the SW. Subsurface soil texture was tightly coupled with precipitation seasonality in accounting for differences in long-term average seasonal aboveground productivity in the Mojave and Sonoran Deserts. The results of this study indicate that the subsurface is a significant factor in modulating aboveground primary productivity, and needs to be included in future modeling exercises of dryland ecosystem response to climate forcing and change.

Canada’s forests are believed to contain 86 gigatons of carbon, stored above and below ground. These forests are large in area, making them difficult to monitor using conventional means. Understanding the carbon cycle and the role of forests as carbon sinks is crucial in the investigation and mitigation of climate change to address national obligations. One economical solution for monitoring the carbon content of Canada’s forests is the development of an automated computer system which uses multisource remotely sensed data to estimate the aboveground carbon of trees. The process involves data fusion of remotely sensed hyperspectral data for tree species information and lidar (light detection and ranging) and radar (radio detection and ranging) for tree height. The size and dimensionality of the data necessitate the efficient use of computing resources for analysis. The outcome is a useful carbon measuring system. The three research questions are: (1) How do we map with remote sensing aboveground carbon in the forests? (2) How do we determine the accuracies of these aboveground carbon maps? (3) How can an automated system be designed for creating aboveground carbon maps? / Graduate

abstract: Climate change will result not only in changes in the mean state of climate but also on changes in variability. However, most studies of the impact of climate change on ecosystems have focused on the effect of changes in the central tendency. The broadest objective of this thesis was to assess the effects of increased interannual precipitation variation on ecosystem functioning in grasslands. In order to address this objective, I used a combination of field experimentation and data synthesis. Precipitation manipulations on the field experiments were carried out using an automated rainfall manipulation system developed as part of this dissertation. Aboveground net primary production responses were monitored during five years. Increased precipitation coefficient of variation decreased primary production regardless of the effect of precipitation amount. Perennial-grass productivity significantly decreased while shrub productivity increased as a result of enhanced precipitation variance. Most interesting is that the effect of precipitation variability increased through time highlighting the existence of temporal lags in ecosystem response. Further, I investigated the effect of precipitation variation on functional diversity on the same experiment and found a positive response of diversity to increased interannual precipitation variance. Functional evenness showed a similar response resulting from large changes in plant-functional type relative abundance including decreased grass and increased shrub cover while functional richness showed non-significant response. Increased functional diversity ameliorated the direct negative effects of precipitation variation on ecosystem ANPP but did not control ecosystem stability where indirect effects through the dominant plant-functional type determined ecosystem stability. Analyses of 80 long-term data sets, where I aggregated annual productivity and precipitation data into five-year temporal windows, showed that precipitation variance had a significant effect on aboveground net primary production that is modulated by mean precipitation. Productivity increased with precipitation variation at sites where mean annual precipitation is less than 339 mm but decreased at sites where precipitation is higher than 339 mm. Mechanisms proposed to explain patterns include: differential ANPP response to precipitation among sites, contrasting legacy effects and soil water distribution. Finally, increased precipitation variance may impact global grasslands affecting plant-functional types in different ways that may lead to state changes, increased erosion and decreased stability that can in turn limit the services provided by these valuable ecosystems. / Dissertation/Thesis / Doctoral Dissertation Biology 2014

The MSc thesis is part of the project of Ministry of Environment of the Czech Republic entitled Czech Terra, which aims at assessing the carbon budget and cycle in the main types of ecosystems in the Czech Republic. This thesis is focussed on the production of aboveground biomass of Carex acuta, which dominates the unmanaged and permanently flooded part of the Wet Meadows. The seasonal dynamics of aboveground plant production was followed using a series of 9 destructive harvests during the vegetation season. On each date, four 0.5x0.5m2 samples were taken from the wetter, and four samples from the drier part of the stand. The maximum value of live biomass of C. acuta (550.8 g.m-2) was recorded on 13.6 2008. The highest value of live biomass of all species reached 602.4 g.m-2. The maximum total biomass (without litter) reached 994.6 g. m-2. The highest value of productivity of C. acuta (12.46 g.m-2.day-1) was recorded on 24. 5.

Metodologias que sejam padronizadas e consolidadas para quantificar carbono em florestas tropicais vem sendo discutidas em convenções climáticas. Este trabalho contribui para a estimativa de biomassa e carbono estocado na parte aérea de áreas de restauração da Mata Atlântica em torno de reservatórios localizados no Pontal do Paranapanema utilizando tecnologia LiDAR (Light Detection and Ranging). Procurou-se explorar o acúmulo e o estoque de carbono em três florestas em diferentes condições de sucessão: duas florestas caracterizdas como madura e a floresta restaurada. Na primeira etapa do trabalho, houve a escolha das equações alométricas encontradas em literatura para determinar a quantidade de carbono em cada uma das áreas utilizando variáveis medidas em campo. Assim, foi utilizada a equação de Ferez et al. (2015), ajustada em floresta restaurada da Mata Atlântica, para quantificar o carbono da área de estudo. Na segunda etapa do trabalho, procurou-se estimar a taxa de incremento do estoque de carbono da parte aérea, utilizando as variáveis medidas em campo obtidas em duas campanhas, nos anos de 2015 e 2016. O corredor restaurado apresentou média de 7,1 Mg.C.ha-1, a floresta madura da ESEC apresentou 39,9 Mg.C.ha-1, e a floresta madura do Morro do Diabo apresentou 45,2 Mg.C.ha-1 para o ano de 2016. A fixação anual encontrada na variação do estoque de 2015 para 2016 foi de 1,2 Mg.ha-1.ano-1 para a floresta restaurada, 1,6 Mg.ha-1.ano-1 para a floresta madura da ESEC e 2.5 Mg.ha-1.ano-1 para a floresta madura do MD. A terceira etapa do trabalho traz a modelagem realizada com dados LiDAR e dados do inventário convencional. Após utilizados métodos estatísticos para seleção de modelo, coeficiente de determinação ajustado, critério de Akaike e erro padrão da estimativa, o modelo escolhido utiliza as métricas percentil 90 e porcentagem de retornos acima de 50cm do solo para estimar carbono acima do solo, obtendo como coeficiente de determinação 0,78. A extrapolação para a área total pelo método do inventário convencional e pelo método da modelagem LiDAR apresentaram diferenças, demonstrando a utilização do LiDAR para reconhecer e retirar informações de áreas não amostradas. O quarto capítulo dessa dissertação traz a variação de carbono estimado pela modelagem com métricas LiDAR para os anos de 2015 e 2016, além do mapa do estocagem evidenciando os locais de maior sucesso e os de menor. A tecnologia LiDAR se mostrou eficiente em captar a variação ocorrida no intervalo de um ano e na quantificação de carbono. / The establishment of methodologies used to quantify carbon in tropical forests are one of the main topics on climate conventions. This project contributes to the estimation of biomass and aboveground carbon stored around reservoirs located in Pontal do Paranapanema, São Paulo - Brazil, using Light Detection and Ranging (LiDAR) technology. The objective was to explore the accumulation of aboveground carbon stored in three different succession conditions: mature, secondary and restored forest. The first chapter of this thesis shows the common allometric equations found in literature used to determine the amount of carbon in each area using field variables. However, the chosen allometric equation was developed by Ferez et al. (2015), because it was adjusted in similar areas than the area of this work. Thus, the restored corridor presented a mean of 7.1 Mg.C.ha-1, the secondary forest 39.9 Mg.C.ha-1 and the mature forest 45.2 Mg.C.ha-1 for the year of 2016. The annual fixation found was about 1.2 Mg.ha-1 for the restored forest, 1.6 Mg.ha-1 for the secondary forest and 2.5 Mg.ha-1 for the mature one. The second article brings the modeling performed with LiDAR data and traditional field inventory data. After using statistical methods for model selection, the chosen model uses two metrics: percentile 90 and percentage of returns above 50 cm of height to estimate aboveground carbon The extrapolation to the total area by traditional inventory and LiDAR modeling showed differences, demonstrating the efficiency of LiDAR to recognize information from non-sampled areas. The last chapter of the thesis brings the variation of carbon using LiDAR data through the years of 2015 and 2016. LiDAR data showed to be useful for measuring aboveground carbon and to detect the increment of the carbon stock over a year.

The ability to accurately assess forest carbon storage is critical to understanding global carbon cycles and the effects of changes in land cover on ecological processes. However, existing methods for calculating carbon storage do not explicitly account for differences in carbon stored by different species of trees. Those methods that do reflect some of this variability, such as remotely-sensing canopy structure to estimate biomass, can be resource-intensive and difficult to reproduce over past or future time steps in order to assess change. I examined the accuracy of several carbon mapping approaches to understand how specificity of forest type classification (for example, classifying forest as "sugar maple/birch" versus simply "deciduous") affects landscape estimates of forest carbon storage in the northeastern United States. I constructed three distinct models to estimate aboveground and coarse roots forest carbon across the study region. These models varied primarily in the specificity of forest type classifications in the input maps and the corresponding carbon storage estimates used for each type. The forest classification schemes tested, from highest to lowest specificity, were: 1) relative basal area by species, 2) species association classes, and 3) coarse forest types (in accordance with IPCC (2006) guidelines). The specificity of forest type classifications in the input maps did influence results, with higher carbon storage estimates generated by models using coarser forest classifications. Maps generated by models that included relative basal area or species association classifications had similar means and standard deviations to the validation plots, as well as the highest correlations with 1000 random points from a remotely-sensed biomass map, suggesting that they better represent variability in carbon storage across the region; however, this variability was largely driven by the incorporation of stand age. Error increased at higher elevations, and decreased with higher total maple-beech-birch components. This likely reflects the dominance of low elevation hardwoods in the studies on which carbon storage estimate tables are based and demonstrates the importance of matching input estimates to region-specific studies. Current estimates of forest carbon storage from the US Forest Service predict 84-90 Mg/ha in this study area, a low value when compared with my modeled estimates of 104 Mg/ha, 108 Mg/ha, and 118 Mg/ha from the relative basal area, species association, and high IPCC models, respectively. If IPCC carbon estimates are to be applied in the northeastern US, the high end of these ranges should be used. Carbon storage estimates that consider different carbon storage capacities of different tree species are useful to explore temporal trends and relative spatial patterns in carbon storage across heterogeneous landscapes, but because of the coarse resolution and low accuracy of existing stand age maps, remotely-sensed biomass maps may be a better approach to quantify carbon stored at specific locations.

京都大学 / 0048 / 新制・課程博士 / 博士(農学) / 甲第21807号 / 農博第2320号 / 新制||農||1065(附属図書館) / 学位論文||H31||N5179(農学部図書室) / 京都大学大学院農学研究科森林科学専攻 / (主査)教授 大澤 晃, 教授 北島 薫, 教授 神﨑 護 / 学位規則第4条第1項該当 / Doctor of Agricultural Science / Kyoto University / DFAM

Thesis (M.Sc. (Geography)) -- University of Limpopo, 2020 / Wetland vegetation provides a variety of goods and services such as carbon sequestration, flood control, climate regulation, filtering contamination, improve and maintain water quality, ecological functioning. However, changes in land cover and uses, overgrazing and environmental changes have resulted in the transformation of the wetland ecosystem. So far, a lot of focus has been biased towards large wetlands neglecting wetlands at a local scale. Smaller wetlands continue to receive massive degradation by the surrounding communities.Therefore, this study seeks to assess and map wetland vegetation as an indicator of ecological productivity on a small scale. The Sentinel-2 MSI image was used to map wetland plant species diversity and above-ground biomass (AGB). Four key diversity indices; the Shannon Wiener (H), Simpson (D), Pielou (J), and Species richness (S) were used to measure species diversity. A multilinear regression technique was applied to establish the relationship between remotely sensed data and diversity indices and AGB. The results indicated that Simpson (D) has a high relationship with combined vegetation indices and spectral band, yielding the highest accuracy when compared to other diversity indices. For example, an R² of 0.75, and the RMSE of 0.08 and AIC of -191.6 were observed. Further, vegetation AGB was estimated with high accuracy of an R² of 0.65, the RMSE 29.02, and AIC of 280.21. These results indicate that Maungani wetland has high species abundance largely dominated by one species (Cyperus latifidius) and highly productive. The findings of this work underscore the relevance of remotely sensed to estimate and monitor wetland plant species diversity with high accuracy.