Influence of the Sea Surface Temperature Decadal Variability on Tropical Precipitation: West African and South American Monsoon by Julián Villamayor

Author:Julián Villamayor
Language: eng
Format: epub
ISBN: 9783030203276
Publisher: Springer International Publishing

Fig. 3.10Correlation (R) between the GW time series and the Sahel index of the JAS precipitation anomalies, 13-year low-pass filtered, from the RCP8.5 future projection of the CMIP5 models individually. Green, yellow and red colors indicate the correlation values that are significant at 95%, 90% or lower confidence level, respectively, following a “random-phase” test

3.2.1 Inter-model Analysis

The historical simulation of the CMIP5 models reproduce, on average, GW time series similar to the observed one. They also simulate broadly the main features of the observed SSTA pattern and the same impact on Sahel rainfall. But individually, some models show important differences with respect to the others. The GW indices of some models have revealed certain deviation with respect to the model-mean, which is also reflected in their individual SSTA patterns (see Appendix A). Consistently, the precipitation response to the GW is not the same in all models.

In order to show the relationship between GW and the Sahel rainfall variability that the CMIP5 models individually reproduce, the correlation between the GW time series and the index of the Sahel rainfall anomalies in JAS, 13-year low-pass filtered, for each model is represented together with observations in Fig. 3.9. Regarding observations, such a relationship is negative, robustly shown by all precipitation data bases (GPCC v7, CRU TS3.24.01 and UDEL v4.01), though it is not statistically significant (with confidence level lower than 70%). With the exception of HadGEM2-CC, HadGEM2-ES, MIROC5, MPI-ESM-LR and MRI-CGCM3, models reproduce a negative relationship in accordance with observations, though not statistically sign for most. Despite this, the general agreement among models and observations supports the confidence in the result that the global warming contributed to a Sahel drying along the 20th century.

Conversely, models show stronger discrepancy regarding the simulated link between the GW and the Sahel rainfalll in the RCP8.5 future projections (Fig. 3.10). 7 out of the 17 CMIP5 models analyzed (CCSM4, CNRM-CM5, FGOALS-g2, MIROC5, MIROC-ESM-CHEM, MRI-CGCM3 and NorESM1-M) simulate anomalous precipitation abundance throughout the Sahel region associated with the GW pattern projected for the future. The rest, instead, reproduce Sahel drying. In general, the correlation between the GW and Sahel indices from the models in this case is higher than in the historical simulation and statistically significant. This suggests that the Sahel rainfall low-frequency variability may be more influenced by the GW in an hypothetical future with high GHGs emission than it was in the 20th century. However, due to the discrepancy among models, the GW effect on the Sahel rainfall projected for the future is largely uncertain. This result is in agreement with other works addressing the future evolution of the Sahel rainfall and showing that the future projections are highly uncertain since GCMs simulate different precipitation trends for the 21st century (Biasutti 2013; Park et al. 2015; Gaetani et al. 2017; Monerie et al. 2017; Biasutti and Giannini 2006; Giannini 2010; Cook and Vizy 2006).

Fig. 3.11Correlation between the regression coefficients of precipitation anomaly averaged over the Sahel (delimited area in Fig. 3.7a) and the regression patterns of GW of the 17 CMIP5 in series in the historical (a) and RCP8.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.