Soil erosion prediction using the Revised Universal Soil Loss Equation (RUSLE) in Google Earth Engine (GEE) cloud-based platform

S Papaiordanidis,I.Z Gitas,T Katagis

doi:10.19047/0136-1694-2019-100-36-52

S Papaiordanidis, I.Z Gitas + Show 1 more

Open Access

https://doi.org/10.19047/0136-1694-2019-100-36-52

Copy DOI

Abstract

High-quality soils are an important resource affecting the quality of life of human societies, as well as terrestrial ecosystems in general. Thus, soil erosion and soil loss are a serious issue that should be managed, in order to conserve both artificial and natural ecosystems. Predicting soil erosion has been a challenge for many years. Traditional field measurements are accurate, but they cannot be applied to large areas easily because of their high cost in time and resources. The last decade, satellite remote sensing and predictive models have been widely used by scientists to predict soil erosion in large areas with cost-efficient methods and techniques. One of those techniques is the Revised Universal Soil Loss Equation (RUSLE). RUSLE uses satellite imagery, as well as precipitation and soil data from other sources to predict the soil erosion per hectare in tons, in a given instant of time. Data acquisition for these data-demanding methods has always been a problem, especially for scientists working with large and diverse datasets. Newly emerged online technologies like Google Earth Engine (GEE) have given access to petabytes of data on demand, alongside high processing power to process them. In this paper we investigated seasonal spatiotemporal changes of soil erosion with the use of RUSLE implemented within GEE, for Pindos mountain range in Greece. In addition, we estimated the correlation between the seasonal components of RUSLE (precipitation and vegetation) and mean RUSLE values.

Highlights

Soil erosion is one of the many natural processes that take place in ecosystems; accelerated soil erosion has a negative impact on agriculture and silviculture, hydrological systems, land degradation, and loss of non-renewable soil resources (Lal, 1998; Morgan, 2009)
The satellite data that we used in this study were four images that were derived from the Sentinel-2 MultiSpectral Instrument (MSI), and were acquired in January 29, April 9, July 3, and October 26, all during 2018
As shown in the results below, Revised Universal Soil Loss Equation (RUSLE) appears to vary through the year

Summary

Introduction

Soil erosion is one of the many natural processes that take place in ecosystems; accelerated soil erosion has a negative impact on agriculture and silviculture, hydrological systems, land degradation, and loss of non-renewable soil resources (Lal, 1998; Morgan, 2009). Traditional field measurements of soil erosion, despite being accurate and reliable, are very expensive and time consuming (Castillo et al, 2012), many scientists turned to predictive models that use satellite data to calculate soil erosion (Wischmeier and Smith, 1978; Lane et al, 2003; Pandey et al, 2007; Rahman et al, 2009). One of these methods is the Universal Soil Loss Equation (USLE), and its descendant, the Revised Universal Soil Loss Equation (RUSLE) (Renard et al, 1991). RUSLE uses multispectral satellite images, as well as satellite-acquire elevation models of the terrain, along with precipitation and soil data (Renard et al, 1997)

Objectives

Methods

Results

Conclusion