Integration of Remote Sensing, GIS and hydrologic Models for predicting Land cover change impacts on watershed runoff and sediment yield

Revision as of 00:59, 5 February 2012 by Mqmallonga (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Jojene R. Santillan

Thesis (M.S. Remote Sensing)--University of the Philippines, Diliman-2008


Land-cover modifications due to changes in land-use have great impacts to natural ecosystems and presently remain a great challenge to environmentalists, watershed planners, policy makers and researchers throughout the world. In the Philippines, the high degree of seriousness and complexities of these impacts to watershed runoff and sediment yield have been exemplified in the Taguibo Watershed in the province of Agusan del Norte in the island of Mindanao. In order to overcome the complexities, to properly understand and assess the impacts, and to formulate proper mitigation measures and efficient conservation and rehabilitation strategies, this study developed and applied a methodological framework that integrated remote sensing (RS) and geographic information system (GIS) with hydrologic and sediment yield models. The US Soil Conservation Service-Curve Number (SCS-CN) based hydrologic model and the modified Universal Soil Equation(MUSLE)-based sediment yield model selected for integration were structured upon a common GIS platform that facilitated data exchange. The capability of remote sensing in providing spatially continuous and multi-temporal land-cover information was utilized in obtaining the land-cover parameters of the model. In particular, Landsat ETM+ and MSS images were analyzed to obtain land-cover maps needed for land-cover parameterizations of the models. Computed values of model performance evaluation statistics such as the Nash-Sutcliffe Coefficient of model Efficiency E, percent bias (PBIAS) and RMSE-Standard Deviation ratio (RSR) indicated that the models have more than satisfactory overall performances. Three scenarios of land-cover change were considered in testing the applicability of the RS-GIS-based hydrologic and sediment yield models for the prediction of land-cover change impacts. The first scenario showed the present land-cover distribution in the study area, with the land-cover map approximated from the Landsat ETM+ image the condition of the study are in 1976 was considered for the second scenario where a land-cover map was obtained from a Landsat MSS image in the third scenario, rehabilitation strategies (such as reforestation of grasslands and agro-forestation of bare soil) that seeks to reduce the impacts of land-cover change to runoff and sediment yield were integrated in the models by imposing the strategies in the present land-cover map. By appropriately changing the land-cover parameters to reflect the changes, the models were run for every scenario to predict watershed runoff and sediment yield resulting from the same rainfall events. Results of model predictions showed that the present condition of the study area is the most vulnerable to the generation of huge amount of runoff and sediment yield. Compared to the present scenario, an 11% reduction in runoff and 31% reduction in sediment yield were predicted by the models when the land-cover condition is that of 1976. The predictions for the rehabilitated state of the watershed signified that the implementation of the rehabilitation strategies will likely reduce the amount of runoff and sediment yield in the present condition by 24% and 96%, respectively.

The results of the integration process demonstrated the synergy that can be attained through the linkage of the hydrologic and sediment yield models. The ability of the framework to quantifiably predict the potential hydrologic implications of land-cover change offers watershed planners and decision-makers a valuable tool for indentifying which proposed land-cover rehabilitation strategies will be effective at minimizing runoff and sediment yield during rainfall events in watershed ecosystems.

Subject Index : Watersheds, Runoff, Remote Sensing