Four generations of the NCAR Community Atmosphere Model (CAM3, 4, 5, and 6) are used to assess model upgrades to the Indian summer monsoon rainfall (ISMR) simulations. The total precipitation simulation in CAM3 is significantly overestimated over the western equatorial Indian Ocean, Arabian Sea, and southwest Bay of Bengal (BoB), and underestimated over the northern BoB due to overestimated convective and underestimated large-scale precipitation, respectively. In the subsequent CAM versions, the simulation of total precipitation has improved, except for the wet bias over the Western Ghats and Himalayan foothills, which has deteriorated due to increased convective and large-scale precipitation, respectively. The improvement in total precipitation simulation over the northern BoB is found in successive CAM versions from the increased heating and drying of the troposphere. CAM3 overestimates the frequency of low precipitation rates and underestimates the frequency of high precipitation rates, which is improved in CAM4, but deteriorated further in subsequent CAM versions. The model development cycle from CAM3 to CAM6 also led to a substantial improvement in most ISMR-associated circulation features, except for the successive overestimation in the low-level monsoon jet due to a strengthened easterly shear of zonal wind. The simulation of subtropical westerly and tropical easterly jets has become more reliable in subsequent CAM versions. In addition, we find successive improvements in the monsoon intra-seasonal oscillation (MISO), associated internal dynamics, and the east-west and north-south heat source. However, some important biases (e.g., the eastward component of MISO, monsoon low-level jet, excessive precipitation over Himalayan foothills, early monsoon onset) need to be alleviated for more realistic ISM simulations by improving further the cloud microphysical and moist processes in the future CAM versions.
Bibliographical noteKAUST Repository Item: Exported on 2023-07-19
Acknowledgements: This work is partly supported by the DST Centre of Excellence in Climate Modeling at IIT Delhi and through a DST Science and Engineering Research Board (SERB) project (ECR/2015/000229). Ravi Kumar acknowledges the MTech fellowship from MHRD. The authors wish to acknowledge NCAR’s climate modeling section (http://www.cgd.ucar.edu/) for developing CAM3, CAM4, CAM5, and CAM6 models. We are thankful to TRMM (https://pmm.nasa.gov/trmm) and ECMWF (https://www.ecmwf.int/en/forecasts/datasets/) for providing the necessary observational and reanalysis datasets publicly. The IITD supercomputing facility is used for computations. NCAR-NCL is used for data analysis, and the climate data operator is used for data processing.
ASJC Scopus subject areas
- Atmospheric Science