Quantification of model uncertainty in the projection of sub-daily maximum wet spell length under RCP 8.5 climate change scenario 

Abstract

<p>Global precipitation characteristics have been significantly altered due to the global warming. While, this is well-known, the sub-daily extreme precipitation events are more sensitive, as compared to the daily-scale. The future intensification of these sub-daily extremes worsen the risk of floods and droughts, thereby posing threat to the natural ecosystem and human society. The ability of general circulation models (GCMs) in simulating the sub-daily precipitation may be inferior, due to their coarser resolutions and complex parametrization schemes. In addition, the characteristics such as the intensity, frequency and duration of sub-daily precipitation may not be correctly simulated by the GCMs. Despite this fact, there are limited studies to investigate the credibility of sub-daily precipitation projections by GCMs, and the related uncertainty. Therefore, in order to investigate the reliability of GCMs in the projections of such extremes, we have used 20 Coupled Model Intercomparison Project phase 5 (CMIP5) models under RCP8.5 (Representative Concentration Pathway). The uncertainty is estimated in the projections of maximum wet spell length (WSL) i.e. maximum number of consecutive wet hours in four different meteorological seasons (DJF, MAM, JJA, and SON), for both near (2026-45) and far future (2081-99) time periods. The equatorial regions of Africa and South East Asia, showed higher model disagreement during every season. In contrast the equatorial regions of South America and South Asia showed significantly more disagreement during DJF and JJA season. Model uncertainty in each hemisphere is observed to be higher during their respective wet seasons. Though the model uncertainty in far future is varying when compared with that in near future, the uncertainty is not increasing globally. Also, the uncertainty is observed to have significantly decreased during MAM season in far future. The spatial contribution towards higher model uncertainty range, is less as compared to lower uncertainty range over the globe. While the magnitude of model uncertainty is varying with time, the latitudinal heterogeneity remains same in both the time period. </p><p>Keywords: precipitation extremes, sub-daily, wet spell, GCM, projections, uncertainty, RCP 8.5</p><p> </p>