Dissertation Neeff
Dissertation
| Modeling tropical forest structure and carbon budget: The Tapajos forest in Central Amazoniavon Till Neeff, 2005 The dynamics of forests growing in Amazonia are vibrant and the structure of vegetation is the most diverse. Ecological processes generate ecosystems that are heterogeneous in space and time. Also, land use dynamics are vigorous. Anthropogenic expansion exerts considerable pressure on forests and results in a patchwork of vegetation types and successional phases. Forest inventory data together with dual-frequency interferometric radar results were used to explore the processes driving heterogeneity and changes in the Tapajos forest in Central Amazonia. A suite of modeling techniques provided means for quantitative description of forests from different viewpoints: statically by standard statistics, across time by change matrices and differential equations, and in space by harmonic analysis and Markov point processes. Forest structure was comprehensively measured from radar imagery providing not only information on all standard forest parameters with excellent precision but also deriving stand tables and classification by successional phases. Spatial patterns in primary forest and among its very large trees were addressed by two different methodological avenues. The consequences of forest dynamics for the regional carbon budget were quantified spatially taking into account the four major vectors of forest modification, and growth dynamics as a function of successional stages. The combination of advanced radar technology with forest inventory results revealed great potential for tropical forest measurement. Great emphasis was allocated to establishing models with a a conceptually convincing rather than merely empirical foundation, which would thus refrain from both oversimplification and excessive complexity in approaching forest processes and their own diversity and heterogeneity. Keywords: Amazon, carbon budget, dual-frequency SAR, ecological modeling, forest structure |
Winter Semester 2024/2025 “Mixed effect models with R”