Session: Cross-cutting Group 2 (Paleovar, Past to future, Data assimilation)
Author: Pascale Braconnot / email@example.com / LSCE/IPSL,unite mixte CEA-CNRS-UVSQ, Université Paris Saclay, orme des merisers, bât. 712, 91191 Gif sur Yvette Cedex
Co-author: Weipeng Zheng, LASG/IAP, Chinese Academy of Sciences, 100029, BeijingBeijing;
The representation of future precipitation changes in West Africa show little agreement between model simulations. In addition, PMIP mid-Holocene simulation underestimate the northward extend of the monsoon rain. Here we propose a convective regime sorted analyses to estimate the relative part of the changes in boreal summer precipitation resulting from changes in large scale pattern or more local thermodynamically factors. For this we consider three periods: Last Glacial Maximum (LGM), mid-Holocene (MH) and abrupt4xCO2 (4X). The first part of the analyses shows that there is more analogy between model results when using this classification, compared to ensemble model mean. It tells us that part of the model spread in the change in total precipitation results from differences in the relative effect of large scale versus more local thermodynamics. Model agreement is however dependent on the analyses period, the largest agreement being found for the changes in the distribution of the convective regimes. In order to show detailed regional changes in WAM precipitation, an attempt is made to modify the precipitation regime of the ERA-interim reanalysis data by the changes of regimes from model results, and to apply a correction on precipitation that takes into account the information gained from the distribution and the efficiency of precipitation of the different convective regimes. For 4X and LGM it is however difficult fully assess regional changes given the regional spread of the results.