GEOPROCESSING TECHNIQUES AND MODELLING APPLIED TO MONITOR EROSION OF TROPICAL
SOILS
Mario Valerio Filho
Geraldo Jose Lucatelli Araujo Junior
Instituto Nacional de Pesquisas Espaciais - INPE,
P.O Box 515, 12.227-010 - Säo José dos Campos - SP-Brazil
Comission VII, Working Group 6
KEY WORDS: Remote Sensing, Modelling, Geoprocessing, Soil Erosion.
ABSTRACT:
This work presents a study that was made in an agricultural watershed (eastern Säo Paulo State, Brazil) and its purpose
was to monitor soil erosion susceptibility areas and to provide information for rural planning. A Geographic
Information System - GIS (SGI/INPE) was used to integrate physical parameters of the Universal Soil Loss Equation -
USLE model adjusted. Erosivity, Erodibility, Slope Length, Slope Steepness, Land Use and Management and
Conservationist Practices USLE factors were integrated in a GIS environment to obtain the soil loss estimation and
characterization of areas under erosion process during 1988 and 1994. The erosion susceptibility maps provide an
estimate and spatial distribution of areas with erosion risk as well as the monitoring of critical areas and provided
further information for soil conservation planning.
1. INTRODUCTION
The great demand for food production for the growing
population, has consequently caused the occupation of all
agricultural lands without considering the capability
potential of these areas. The results of this occupation is
the loss of natural soil fertility, changes of water quality
and silting of rivers and reservoirs.
For a rational land ocupation and agricultural production
it is important to establish agricultural policies which
consider the food production needs and environmental
conservation issues, specially considering soil and water
conditions.
Geoprocessing techniques envolving remote sensing and
geographic information systems are important tools
which allow the acquisition and integration of thematic
data (Bocco and Valenzuela, 1988; Zhou, 1989; Castro,
1992; Olson et al., 1994) and also provide inputs for the
spatial/temporal characterization of areas under erosion
process at the watershed level, and characterization of
soil erosion susceptibility.
In this context several works have been developed using
loss estimate model Universal Soil Loss Equation-USLE
(Wischmeier and Smith, 1978). There are several studies
using USLE model such as: Ventura et al. (1988); Pinto
(1991); Donzeli et al. (1992); Valerio Filho et al. (1993);
and Mellerowicz et al. (1994).
This study presents a methodology of modelling and
geoprocessing techniques for the characterization and
237
monitoring of areas submited to soil erosion in 1988 and
1994.
2. METHODOLOGICAL PROCEDURES
The area under study is the watershed of Ribeiräo Bonito,
located around the lake of Barra Bonita, in the centre of
Sáo Paulo State, Brazil, located at S 22°38’ 11°, to
S 22? 2T 15” and W 42° 11° 52” to W 42° 06’ 16”.
The Ribeiräo Bonito watershed is a natural landscape unit
insert among the boundary of two larger geologic-
geomorphologic units: Peripheral Depression, made up
mainly by sandstones and the Front of Cuestas, a
structure of sandstones and basaltic rocks. The
escarpment, where this river is born, defines different
runoff conditions, mainly at the medium and lower
course at the Periferic Depression, characterizing wide
and smooth interfluves and low relief energy.
The soil map is represented by six large groups with
variations in physical and chemical characteristics.
Agricultural production is concentrated at the lower and
middle portion of the wathershed, represented mainly by
sugar cane plantations, summer crops, pastures and
reforestation.
To perform this study, the USLE - Universal Soil Loss
Equation, (Wischmeier and Smith, 1978) was used as an
analytical guide to process and integrate the
environmental data available. The USLE model is a
function of six environmental and antrophic variables:
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996
