University of Twente Student Theses


Analysis of the effect of up-scaling and soil erosion assessment at regional scale by using Pesera model: A case study of Lomsak District, Phetchabun, Thailand

Shrestha, Bharat Babu (2009) Analysis of the effect of up-scaling and soil erosion assessment at regional scale by using Pesera model: A case study of Lomsak District, Phetchabun, Thailand.

[img] PDF
Abstract:Erosion assessment at regional scale in heterogeneous and fragmented land use/land cover situation of Lomsak district of Thailand is an urgent need to address the increasing erosion problem. Therefore PESERA, a physically based and spatially distributed regional scale model was used to assess soil erosion in Lomsak district of Thailand. The study has also focussed to see the effect of up-scaling of grid cells of land use/land cover on erosion prediction. The model was run by using 128 data layers related to climate, soil, land cover and topography factors. Climate parameter maps were prepared by interpolating climate data considering the situation of climatic variation with altitude. Soil parameter maps were generated by ordinary kriging interpolation method on the basis of analyzed soil data collected randomly from the study area. For preparing land use/land cover map Aster image of Feb. 4, 2006 was classified by supervised classification with nearest neighbour algorithm. Monthly canopy cover maps prepared from both the crop calendar and NDVI data were used separately for their comparative study. DEM as a topographic parameter was prepared by interpolating from contour maps of 20m interval. Model was run using 250m, 500m and 1000m resolution data to see the effect of up-scaling. The model result shows that the rate of soil erosion spatially varies from < 1.0 ton/ha/year to 19.2 ton/ha/year. There is not much difference in average annual soil erosion rate between model results whether crop calendar (1.17 ton/ha/year) or NDVI (1.13 ton/ha/year) data is used. The soil erosion rate has also varied with the seasonal and land use/land cover variation. The highest erosion rate occurs during month of June and followed by August and October having second and third highest erosion rate respectively. Agriculture land generates more soil erosion than natural land. Maize field has highest erosion rate while dense forest and grass land have the least erosion rate. Erosion assessment at various land use scenarios shows that the conversion of land use from degraded forest and grass land to agriculture has resulted substantial increase in erosion rate. Erodibility is the most sensitive model parameter among the selected parameter used for sensitivity analysis. There is not difference in average annual soil erosion rates of the whole study area whether 250m, or 500m data is used, however, there is significant difference in total erosion with respect to various land use classes as there is change in areas of land use/land cover classes while up scaling the grid cells of land use/land cover map. Key words: PESERA model, up-scaling, NDVI, crop calendar, canopy cover, land use scenarios, DEMs.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page