Possible changes in Sudan's future precipitation under the high and medium emission scenarios based on bias adjusted GCMs

Abstract

Prospecting our future climate behavior under a wide range of possible scenarios is one of the most challenging issues in climate science. This study examines changes in monsoon's (June–September) future precipitation over three zones distributed across Sudan based on GCMs from CMIP5 and (CMIP6). The models are GISS-E2-H, IPSL-CM5A-MR, and MPI-ESM-LR (BCC-CSM2-MR, INM-CM4-9 MPI-ESM1-2-LR) in addition to the ensemble mean of each group. Given systematic errors in the GCMs simulations, we applied the Quantile Mapping bias correction technique to rectify the bias toward the real-occurred field. The bias coefficients were calculated for 1946–1975 and validated for 1976–2005. The improvement was quantified using the Skills Score technique. Overall, the GCMs simulations improved by 0–15% (45–90%) for CMIP6 (CMIP5) after bias correction (BC) was involved. BC has led to a remarkable breakthrough in the GCMs simulations manifested by a considerable improvement in the simulated precipitation, particularly in southern parts of Sudan. Furthermore, we applied the bias adjustments of the future period 2030–2099 to reduce the uncertainty margins when assessing the possible changes. The future period was divided as (i) Near Future 2025–2049, (ii) Middle Future 2050–2074, and (iii) Far Future 2075–2099. The changes were examined based on two Representative Concentration Pathways (Shared Socioeconomic Pathway) scenarios, namely: RCP4.5(SSP2–4.5) and RCP8.5(SSP5–8.5) for CMIP5 (CMIP6). In both scenarios, June–September is projected to remain the rainy season in all analyzed zones. Among all studied periods for both scenarios, the most severe changes are expected in the far future with a positive change of 60 to >80% shown in CMIP6 projection under the high emission scenario in zone3. However, the highest negative bias is expected in the near future period under the high emission scenario in CMIP5 projections in the same zone. Higher uncertainty is associated with the future projections, particularly in zone3 under the high emission scenario for both CMIP5 and CMIP6 projections. Higher agreement between the models, on the other hand, is shown in zone2 under the high emission scenario for both simulations.