Extratropical Influence on the Sub-seasonal Rainfall Deficit in Indian Summer Monsoon

Abstract

The variability of Indian monsoon on a range of timescales - seasonal to interannual to decadal - has been studied and characterised extensively over the past several decades. On the mechanistic front, the role of tropical oceanic markers, especially the phases of El Nino Southern Oscillation (ENSO), on seasonal rainfall has received justifiably substantial attention. While the footprint of extratropics has been recognised, it has mostly been case-study based, and the overall impact has not been quantified systematically. To this end, the work presented here revisits the influence of extratropics on Indian monsoon: Using a suite of observational and reanalysis data, and idealised numerical experiments, we provide robust evidence of the significant role that midlatitudes play in modulating sub-seasonal variability of monsoon rainfall.

The thesis journey begins with one end of seasonal extremes, namely, droughts. Over the past century, India has experienced droughts on-an-average once in five years (23 in 115 years). Out of these, El Nino is associated with nearly half of them; the others occur when the equatorial Pacific waters are near-neutral. Based on a systematic and exhaustive analysis of observed rainfall, we find that the distinction between these two categories of droughts is further reflected in the evolution of rainfall deficit through the season. In one category, namely, El Nino droughts (EN + Dr), the deficit sets in around mid-to-late June and persists nearly throughout the season. However, in Non El Nino droughts (NEN + Dr), a marginal deficit in early season rainfall is compounded by a dramatic late-August deficit. This sub-seasonal signature differs from the conventional notion of seasonally-persisting rainfall deficit during droughts. In addition, in NEN+Dr years the only global oceanic marker of prominence is a spatially coherent, anomalously cold North Atlantic ocean. Using reanalysis data, we show that the severe and abrupt late-season rainfall deficit can be attributed to a Rossby wave from the midlatitudes. In particular, we find a strong and uncharacteristically persistent deep tropospheric cyclonic vorticity forcing co-located over the cold North Atlantic. The interaction of upper-level winds with this episodic vorticity anomaly results in a wavetrain that curves towards South Asia late in the season, thereby derailing the monsoon.

Following this, we turn our attention to understanding the overall influence of midlatitudes on the sub-seasonal variability of monsoon rainfall. Based on preliminary evidence in the first part of this thesis, we focus on early-season rainfall deficit, with no particular attention to the final seasonal state. Our analysis suggests that over the past century nearly one-third of the years (36) show significant deficit in June rainfall. Out of these, in nearly two-thirds of the years (24), the Pacific SST shows no anomalous behaviour. A composite dynamical analysis of these years reveals that the early-season deficit in these years can also be associated with a Rossby wavetrain from the midlatitudes. Furthermore, using a wave-activity flux approach, we show that the origin of this atmospheric disturbance is the North Atlantic (similar to the late season observation in droughts). Taken together, the midlatitudes appear to have an important role to play in modulating the early- and late-season rainfall.

The motivation for this journey's last leg is the presence of a strong vorticity forcing above an anomalously cold ocean in the North Atlantic. We explore these aspects using two sets of idealised experiments: (i) Specifying an atmospheric vorticity forcing over the North Atlantic in a simple 1-layer shallow-water model; and (ii) Specifying an SST gradient in the North Atlantic basin in an Aquaplanet mode of the Community Earth System Model. The shallow-water model demonstrates similar Rossby wave response to the vorticity forcing over the North Atlantic that could reach over India. The aquaplanet model shows modest success in simulating coherent episodic vorticity forcing over the North Atlantic, in response to an underlying SST gradient between the tropical and northern Atlantic ocean basins. The SST gradient results in a baroclinic instability leading to a strong vertical wind shear which in turn produces a deep tropospheric vorticity over the North Atlantic. This anomalous vorticity forcing then triggers waves which travel towards the Equator (and India), quite similar to what was observed.

The findings of this thesis suggest that the footprint of midlatitudes on Indian monsoon is significantly larger than previously recognised. The atmospheric teleconnection identified here offers a new pathway to understand monsoon rainfall variability on sub-seasonal timescales, especially in the absence of a strong tropical forcing. The broader implication of this work is that prevalent tropics-centred paradigms of Indian monsoon rainfall need to be revisited, and due attention be given to more holistic frameworks that can account explicitly for the influence of extratropics.