Influence of upper ocean on Indian summer monsoon rainfall: studies by observation and NCEP climate forecast system (CFSv2)

Abstract

This study explores the role played by ocean processes in influencing Indian summer monsoon rainfall (ISMR) and compares the observed findings with National Centers for Environmental Prediction (NCEP)-coupled model Climate Forecast System, version 2 (CFSv2). The excess and deficit ISMR clearly brings out the distinct signatures in sea surface height (SSH) anomaly, thermocline and mixed layer depth over north Indian Ocean. CFSv2 is successful in simulating SSH anomalies, especially over Arabian Sea and Bay of Bengal region. CFSv2 captures observed findings of SSH anomalies during flood and drought (e.g., Rossby wave propagation which reaches western Bay of Bengal (BoB) during flood years, Rossby wave propagation which did not reach western BoB during drought). It highlights the ability of CFSv2 to simulate the basic ocean processes which governs the SSH variability. These differences are basically generated by upwelling and downwelling caused by the equatorial and coastal Kelvin and Rossby waves, thereby causing difference in SSH anomaly and thermocline, and subsequently modifying the convection centers, which dictates precipitation over the Indian subcontinent region. Since the observed SSH anomaly and thermal structure show distinct characteristic features with respect to strong and weak ISMR variability, the assimilation of real ocean data in terms of satellite products (like SSHA from AVISO/SARAL) bestow great promise for the future improvement.

Clouds–SST relationship and interannual variability modes of Indian summer monsoon in the context of clouds and SSTs: observational and modelling aspects

“…As a result CFSv2 exhibits pervasive cold SST bias over entire Indian Ocean region (Figure (c)). It is in line with previous studies (Pokhrel et al , ; Saha et al , ; Chaudhari et al , ). In spite of strong cold SST bias in CFSv2 (Figure (c)), rainfall pattern over Bay of Bengal and equatorial Indian Ocean are reasonably good (Figure (c)).…”

“…Therefore, it is a basic requirement that model should simulate reasonable mean monsoon in terms of rainfall, wind circulations, surface temperature, SSTs and so on. Previous studies have (Saha et al , ; Chaudhari et al , , ; De et al , ; Hazra et al , ; Pokhrel et al , ) pointed out that CFSv2 is able to simulate the mean spatial pattern of rainfall, SST, clouds and wind patterns quite realistically. Before delving into the relationship of clouds with SST and rainfall, it will be interesting to explore the mean patterns of clouds, SST and rainfall.…”

Hindcast skill improvement in Climate Forecast System (CFSv2) using modified cloud scheme

“…Despite suitability of CFSv2 for monsoon forecast, the amplitude of mean seasonal monsoon rainfall is greatly underestimated over the monsoon domain (e.g., Pokhrel et al, 2012b;Abhilash et al, 2014;Saha et al, 2014b;Sahai et al, 2015;Chaudhari et al, 2016a). Several diagnostic studies by previous researchers have brought out several prominent bias affecting the rainfall forecast using both free run (e.g., Hazra et al, 2016;Chaudhari et al, 2016c) and hindcast run (e.g., Pokhrel et al, 2016;Saha et al, 2016). Despite considerable model development progress, the modellers are not able to completely answer the huge underestimation of CFSv2 precipitation simulations over the Indian land regions.…”