Long-Term Changes in the Intra-seasonal Modulation of Extreme Rainfall over India

Problem

Extreme monsoon rain does real damage in India, and part of its variability is organised by large tropical waves: the Madden-Julian Oscillation and convectively coupled equatorial waves. Earlier work measured that influence; this thesis asked whether it has changed over four decades.

What I did

Using IMD’s 0.25° daily gridded rainfall for June–September 1979–2018, I defined extremes per grid point as days above the 95th percentile of rainy days, so “extreme” means the same thing in a dry state as on the wet coast. I tracked wave activity with filtered outgoing longwave radiation, split days into wet and dry phases of four wave types, and compared how each phase changed the probability and intensity of extremes in 1979–1998 against 1999–2018, using ERA-Interim 850 hPa humidity and wind composites to look for why.

Result

Wet phases raise the odds of an extreme day and dry phases lower them, but the size of that effect shifted: the depression-type waves’ wet-phase boost fell from about 50% to 31%, while the MJO and equatorial Rossby wet phases grew slightly. Weaker low-level moisture and winds during wet phases go with the reduced modulation. Intensity of extremes changed little outside a few localised areas.

Tools

MATLAB · NCL / GrADS · CDO · NCO · IMD gridded rainfall · NOAA OLR · ERA-Interim