Global rainfall erosivity change projections (2050 RCP 8.5)
The erosive force of rainfall (rainfall erosivity) is a major driver of soil, nutrient losses worldwide and an important input for soil erosion assessments models. This map shows the geographical distribution of erosivity changes for RCP8.5 for the period 2010–2050.
Related Policies and targets
Topics
Data provider
Suggested citation
Panagos, P., Borrelli, P., Matthews, F., Liakos, L., Bezak, N., Diodato, N. and Ballabio, C., 2022. Global rainfall erosivity projections for 2050 and 2070. Journal of Hydrology, Art.no.127865.DOI: 10.1016/j.jhydrol.2022.127865
Related tool or portal
Date or time period of observation
Geographic coverage
Cautions
The global rainfall erosivity projections do not challenge any local or regional studies which have been developed with higher gauge station density and using higher temporal resolution rainfall fata. The number of stations with measured erosivity is rather limited, allowing us only to predict and interpret trends in the future global dynamics of rainfall erosivity rather than deterministically predict exact changes in magnitude. In addition, the confidence given to the spatial patterns of projected rainfall erosivity in this study should be interpreted based on the underlying rain gauge density in the GLoREDa database. It remains difficult to find and collate high temporal resolution (e.g. sub-hourly) rainfall data at a global scale. With 3,625 stations in 63 countries, GloREDa represents the current best effort at compiling the high temporal resolution gauge data necessary for rainfall erosivity calculations. The Global Rainfall Erosivity Database (GloREDa) will become available with the stations values, allowing the further inclusion of data from areas in which we lack baseline information to model future relationships (e.g. Africa). Additionally, using GloREDa as a data tool, future work can help to overcome some of the problems related to spatial and temporal scale when using coarse gridded observations to predict rainfall erosivity which involves fine spatiotemporal characteristics. The use of alternative methods to estimate erosivity such as the Climate Prediction Center MORPHing technique (CMORPH) proved to smoothen areas with high erosivity (Kim et al., 2020), showing that the better accountancy of extremes in continental to global scale modelling remains an important frontier to properly capture to the soil erosion risk associated with rainfall erosivity (Bezak et al., 2022). The Coupled Model Intercomparison Projection (CMIP) has now released the data for the phase 6 while here we have used the data in the phase 5 (CMIP5). Since downscaling is a vital pre-requisite for a study of this kind, we only considered downscaled simulations. Due to the lack of a publicly available CMIP6 dataset with downscaled projections at this time, here we have used the data from the downscaled phase 5 (CMIP5) assessment. An additional limitation comes from the use of WorldClim as input data since the unequal time intervals (2010, 2050 and 2070) doesn’t give an optimal overview of the impact of evolution in the climatic state on rainfall erosivity. In addition, the available (bio)climatic layers represent time slices of 20 years for the 2050 (2041–2060) and 2070 (2061–2080) climatic periods compared to 30 years for the historical baseline period. In climate change studies (Hirabayashi et al., 2013), the 30-year time-slice is most commonly used to represent the average climatic state. However, WorldClim is the only available dataset with historical and future climatic projections that give harmonised covariate inputs matching the methodological approach of this study at a global scale. In case there are no future open-source data releases overcoming these outlined limitations, future global-scale studies may seek to implement their own GCM downscaling regimes to capture the climatic variability at suitable scales for erosion studies. For regional and local assessments, Regional Climate Models (RCMs) provide detailed estimates of meteorological parameters and fill the gap of coarse resolution GCM data (Tapiador et al., 2020).
Select for mapping