Can the introduction of the topographic indices in LPJ-GUESS improve the spatial representation of environmental variables?

Abstract: Ecosystem modelling is an always evolving science trying to catch the complexity of the nature and its principles to model environmental responses in a realistic way. Over and over, models try to introduce more variables and interactions to achieve better representations of phenomena of interest like the responses of the ecosystem to a fast changing world (climate change, land use change). LPJ-GUESS is a flexible dynamic ecosystem model widely used to model the structure and dynamics of terrestrial ecosystems. It is based on plant physiology, biochemical cycles and feedbacks on independent gridcells, there is no consideration of lateral transfer of water between cells. On the other hand, soil moisture is essential for vegetation growth and its distribution is known to be driven by the topography of the landscape, which drives the lateral transfer of water. Based on this, it was considered important to assess the modelled spatial representation of environmental variables (soil moisture, LAI) from LPJ-GUESS and to evaluate a possible method to include the effect of topography over the hydrology in LPJ-GUESS model. For this, Alergaarde catchment (smooth relief) was chosen and by the use of correlation analysis and visual interpretation the following issues were studied, 1) Importance of topography on the spatial distribution of environmental variables based on topographic indices (Ln (Drainage area), tan ( angle slope) and topographic wetness index, TWI); 2) LPJ-GUESS ability to catch the environmental variables spatial distribution and 3) Implementation of a coupled LPJ-GUESS - topographic indices model to account for the topography influence on hydrology and assessment of its performance on modelling the spatial patterns of environmental variables. Results of the first two topics showed how LPJ-GUESS could not catch the spatial variations of satellite based LAI, and that even the gentle topography of the catchment was an important issue on explaining the heterogeneity of vegetation related variables. Nevertheless, there are many factors, like climate conditions, which affect the strength of this relationship, as reflected on low correlation coefficients (never over 0.25), the variable correlation coefficients along the year and the identification of areas more related to the topographic indexes than others. Additionally, TWI was selected, based on its higher correlations with respect to the other topographic indices, to be one used to represent the topography influence in the catchment. The integrated model, LPJ-Topographic index (LPJ-TI), use the TWI to make a cell wise characterization and create weights affecting the water inputs to the soil layer as a way to account for hydrological processes driven by topography. LPJ-TI showed localized and time dependent improvement of the spatial representation of the satellite based LAI. These results confirm the need to include the topographic influence on the hydrological module of LPJ-GUESS and present a possible low computational method to start working on.