Projections of rainfall erosivity in climate change scenarios for the largest watershed within Brazilian territory

Abstract

Global climate change can potentially threaten agricultural production due to endangered natural resources, such as rainfall patterns. Thus, extreme rainfall events can cause greater rainfall erosivity, consequently, greater soil erosion. Conversely, a reduction in rainfall amount can lead to water scarcity for the agriculture production process. This way, it is a foremost need to model climatic conditions under global climate change scenarios, particularly in places where rainfall data tends to increase. This work aimed to project rainfall erosivity in the major Brazilian watershed, the Tocantins-Araguaia river basin, throughout the 21st century under two Intergovernmental Panel for Climate Change Fifth Assessment Report (IPCC AR5) Scenarios, the Representative Concentration Pathways, RCP4.5 and RCP8.5 scenarios. This study uses the downscaling of four global climate models of the Coupled Model Intercomparison Project (CMIP5) by the Eta regional climate model, used by the Brazilian National Institute for Space Research. The average rainfall erosivity was calculated based on the Modified Fournier Index in three periods of 30-year length throughout the 21st century. Time series of R-factor were analyzed at rain gauge station points overlapping regional model grid cells over the basin for the 1961–2099 period. Projections indicated lower annual average rainfall erosivity values in comparison with historical data. Estimated mean rainfall erosivity values were 10,977.69 ± 526 MJ mm ha−1 h−1 yr−1 for the RCP4.5 scenario, and 10,379.71 ± 723 MJ mm ha−1 h−1 yr−1 for the most pessimistic climate change scenario, RCP8.5. The largest reductions of the mean R-factor reached 5,5% for the multi-model ensemble projections for near future, and 15.4% for the ensemble projections models for long-term, with the greatest decreasing trends under RCP8.5. Reductions greater than 2,000 MJ mm ha−1 h−1 are expected throughout the 21st century according to multi-model ensemble projections models under RCP8.5 scenario in most of the watershed. Decreasing rainfall erosivity factor in both RCP scenarios was due to a lower rainfall depth. However, the value of rainfall erosivity is still considered high and should be taken into account in soil conservation practices. Furthermore, the smaller rainfall amount indicates a possible reduction in water availability for crops of longer cycle, and increase in spatial variability of less intense rainfall.