Advanced search
Start date
Betweenand


Evaluation of water erosion prediction models MUSLE and WEPP: contribution in watersheds

Full text
Author(s):
Isabel Cristina Moraes
Total Authors: 1
Document type: Doctoral Thesis
Press: Rio Claro. 2016-04-29.
Institution: Universidade Estadual Paulista (Unesp). Instituto de Geociências e Ciências Exatas. Rio Claro
Defense date:
Advisor: Sergio dos Anjos Ferreira Pinto
Abstract

The land use intensification becomes more and more evident due the growing of food production and supply to the human population. The application of inadequate practices to the multiple functions has shown the accelerated water erosion as a major degradation processes for this resource. In an effort to colaborate for soil conservation studies, this research aimed to evaluate the predictive models of erosion MUSLE – Modified Universal Soil Loss Equation (Williams, 1975) and WEPP – Water Erosion Prediction Project (Laflen, 1991) by comparing the estimated soil loss and empirical data from pins experimental plots. The variability of simulated parameters led the definition of two watersheds with different hydrossedimentological characteristics, watershed of stream Jacutinga (Rio Claro – SP/Brazil) and stream Monjolo Grande (Ipeúna – SP/Brazil) to assess the erosive behavior in soils derived from mudstones and sandstones, respectively. In the experimental period of June/2013 to August/2014, experimental plots of 1 m2 were monitored in different conditions of slope, topographic position, slope shape and vegetation cover - grazing, sugarcane, riparian vegetation and bare soil. In Lixisols and Cambisols clayey the losses registered were 0.018 to 0.066 ton/ha.year-1 and deposition between 0 and 0.026 ton/ha.year-1 , and in Quartzarenic Neosols and Lixisols of sandy texture, where dominated depositional process, between 0.004 and 0.085 ton/ha.year-1 , and the losses were 0.002 to 0.11 ton/ha.year-1 . The greatest losses are associated with the largest portion of fine sand and silt fractions, and weak aggregation by low levels of clay and organic matter. The highest percentage of vegetation cover was crucial for smaller losses, showing greater soil protection factor for direct plantation which in conventional tillage. The simulations were performed for the experimental year and specific rainfall events: Event 1 - intensity greater than 50 mm/h, and Event 2 - intensity less than 14 mm/h. The evaluation of the models was given by the coefficients of determination (R2 ), Pearson (PR2 ) and NashSutcliffe (COE), the latest being the most eficiente one. The MUSLE was simulated obtaining the ramp length (L) by the automatic method (Desmet and Govers, 1996) and manual (linear distance). Estimates were satisfactory to very good (average COE> 0.7) especially in clay soils with predominance of soil losses, and manual factor L. For WEPP, the simulations were performed by hill module (WEPP software) and watershed (GeoWEPP – GIS version for WEPP). In all aspects, WEPP simulations have obtained unsatisfactory results (COE < 0). In GeoWEPP, the estimatives given in loss and deposition classes showed coincidence of classes and records of the plots between 20% and 90% in clayey soils, and between 40% and 50% in sandy soils. The best estimates occurred for low intensity rains, pointing out the inadequacy of this model in extreme events and in tropical conditions. It is emphasized that the rain in the experimental period atypical, with low rainfall compared to historical annual averages. Considering the overestimated values of WEPP and GeoWEPP, only MUSLE was adequate as a tool to water erosion studies, especially for the contribution of this model in land use management, in the planning of agricultural systems and conservation practices. (AU)

FAPESP's process: 12/08710-4 - Performance analysis of models of soil loss MEUPS and WEPP: contribution to implementation in watersheds
Grantee:Isabel Cristina Moraes
Support Opportunities: Scholarships in Brazil - Doctorate