Description: We investigate the effect of variable marine biogeochemical light absorption on Indian Ocean sea surface temperature (SST) and how this affects the South Asian climate. In twin experiments with a regional Earth system model, we found that the average SST is lower over most of the domain when variable marine biogeochemical light absorption is taken into account, compared to the reference experiment with a constant light attenuation coefficient equal to 0.06 m−1. The most significant deviations (more than 1 ∘C) in SST are observed in the monsoon season. A considerable cooling of subsurface layers occurs, and the thermocline shifts upward in the experiment with the activated biogeochemical impact. Also, the phytoplankton primary production becomes higher, especially during periods of winter and summer phytoplankton blooms. The effect of altered SST variability on climate was investigated by coupling the ocean models to a regional atmosphere model. We find the largest effects on the amount of precipitation, particularly during the monsoon season. In the Arabian Sea, the reduction of the transport of humidity across the Equator leads to a reduction of the large-scale precipitation in the eastern part of the basin, reinforcing the reduction of the convective precipitation. In the Bay of Bengal, it increases the large-scale precipitation, countering convective precipitation decline. Thus, the key impacts of including the full biogeochemical coupling with corresponding light attenuation, which in turn depends on variable chlorophyll a concentration, include the enhanced phytoplankton primary production, a shallower thermocline, and decreased SST and water temperature in subsurface layers, with cascading effects upon the model ocean physics which further translates into altered atmosphere dynamics.

Description: An effort is made to implement a regional earth system model (RESM); ROM, over CORDEX-South Asia (SA). The added value of RESM is assessed for mean precipitation, its variability (intraseasonal to interannual), extremes, and associated processes. In this regard, ROM’s fields are compared with the respective fields of its standalone version (REMO), the models belonging coupled model intercomparison project (CMIP5 and CMIP6), and regional climate models of CORDEX-CORE simulations. RESM shows substantial improvement for most of the Indian monsoon’s aspects; however, the magnitude of the value addition varies spatiotemporally and also with different aspects.. The improved representation of intraseasonal variability (active-break spell’s duration and intensity) and Interannual variability attributed to improved mean seasonal precipitation. Additionally, correct representation of sea surface temperature, Indian Ocean Dipole, and its underlying dynamics also contribute to improving the mean precipitation. The notable improvement is seen especially over the south-eastern regions of the Bay of Bengal (BoB) and South-Central India, where increasing (decreasing) low-pressure systems over Central India (BoB) are noticed as a consequence of air-sea coupling, leading to enhanced (reduced) precipitation over Central India (BoB), reducing dry (wet) bias found in REMO and the other models. Despite substantial improvements, RESM has a systematic wet bias in the mean precipitation associated with a warm bias over the western coast of the Arabian Sea. An overestimation of very high extreme precipitation due to the enhanced contribution of low-pressure systems indicates the model’s limitations, suggesting the need for further tuning of the RESM.

Description: High-quality climate information at appropriate spatial and temporal resolution is essential to develop and provide tailored climate services for Africa. A common method to produce regional climate change data is to dynamically downscale global climate projections by means of regional climate models (RCMs). Deficiencies in the representation of the sea surface temperatures (SSTs) in earth system models (ESMs) and missing atmosphere–ocean interactions in RCMs contribute to the precipitation bias. This study analyzes the influence of the regional atmosphere–ocean coupling on simulated precipitation and its characteristics over Africa, and identifies those regions providing an added value using the regionally coupled atmosphere–ocean model ROM. For the analysis, the MPI-ESM-LR historical simulation and emission scenario RCP8.5 were dynamically downscaled with ROM at a spatial resolution of 0.22° × 0.22° for the whole African continent, including the tropical Atlantic and the Southwest Indian Ocean. The results show that reduced SST warm biases in both oceans lead to more realistic simulated precipitation over most coastal regions of Sub-Saharan Africa and over southern Africa to varying degrees depending on the season. In particular, the annual precipitation cycles over the coastal regions of the Atlantic Ocean are closer to observations. Moreover, total precipitation and extreme precipitation indices in the coupled historical simulation are significantly lower and more realistic compared to observations over the majority of the analyzed sub-regions. Finally, atmosphere–ocean coupling can amplify or attenuate climate change signals from precipitation indices or even change their sign in a regional climate projection.