Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Chine → 2022 → Evaluation and Projection of Extreme Precipitation and Drought Using CMIP6 High ResMIP Simulations over West Africa

Nanjing University of Information Science and Technology (2022)

Evaluation and Projection of Extreme Precipitation and Drought Using CMIP6 High ResMIP Simulations over West Africa

Ajibola Felix Olabamiji

Titre : Evaluation and Projection of Extreme Precipitation and Drought Using CMIP6 High ResMIP Simulations over West Africa

Auteur : Ajibola Felix Olabamiji

Grade : Doctoral Dissertation 2022

Université : Nanjing University of Information Science and Technology

Résumé partiel
Climate study over West Africa is a rigorous activity due to topography of the region.The climate models used has to be hi tech simulation and one of the determinant of the model performance is its resolution.This thesis evaluates eight models from High Resolution Multiple-model Intercomparison Project(High Res MIP),a consortium of Couple Model Intercomparison Project Phase 6(CMIP6).The aim is to investigate the performance of improved horizontal(higher)resolution on the corresponding low resolution simulations in reproducing extreme precipitation and drought over West Africa.Projection of extreme precipitation and drought were also performed after models evaluation.The methods employed in this study include calculating the arithmetic average of low and high resolution simulations,precipitation anomaly,Probability Density Function(PDF),Taylor diagram,Pearson correlation coefficient(r),standard deviation(STD),Standardized Precipitation Evapotranspiration Index(SPEI),Precipitation Concentration Index(PCI),Mann Kendall trends analysis,Z-statistics.The study covers the historical period of 1950 – 2014 and 2025 –2049 for projection.Precipitation dataset from Global Precipitation Climatology Centre(GPCC),Climate Research Unit(CRU)TSv4.03,and University Delaware(UDEL)v5.01 were used as reference with the purpose of accounting for uncertainty.The findings are listed below1.The observed precipitation annual peak over Guinea Coast,Sahel and entire West Africa is around 250,150,and 120 mm/m during September,August,and August,respectively.The High Res MIP models overestimates the precipitation onset over the Guinea Coast,but show a high skill over Sahel.The GCMs simulated early annual rainfall peak around May over Guinea coast.The high rainy area from observation is southeast and southwest,the pattern is adequately reproduced by the GCMs,although this is overestimated by some of the models during the monsoon period(June to September).The pattern of observed extreme precipitation over southeast and southwest of the region seems to be adequately reproduced by most of the high and low resolution simulations,although with variations in the simulated amount.However,the Mann Kendall trends analysis for extreme precipitation shows the ability of high resolutions simulations in capturing the significant increasing trends over the two subregions.The observed highest 1-day precipitation seems to be overestimated by majority of models with higher resolution and underestimation by the simulation with low resolution.During June – September(JJAS),CMIP6 simulations showed a high correlation coefficient,low Root Mean Square Error(RMSE)and ratio of standard deviation around 1 for mean climatology and spatial correlation coefficient relatively lower during extreme events.Meanwhile,the orography induced rainfall is adequately reproduce by high resolution models even when observation could not capture it.The improved horizontal resolution does not remove biases in the GCMs,therefore the bias correction employed improve the model performance.2.The projected precipitation annual cycle reveal precipitation increase from 130 to 150mm/m over Sahel while the peak month still remains at August.Over Guinea coast,the pattern of High Res MIP simulation is not consistent with the regular pattern of precipitation over the region(i.e.,the projected annual peak is May which appears to be early)

Présentation (CNKI)

Page publiée le 17 octobre 2022