The effect of irrigation on pre-monsoon rainfall in India

Abstract:

It has been well-established that irrigation is and will continue to be a profound perturbation on the South Asian monsoon system. Widespread irrigation in the Indo-Gangetic plain measurably reduces the surface temperature and surface sensible heat flux over much of the region. General circulation model (GCM) studies have found that irrigation reduces the continental scale land-sea temperature contrast, weakening the monsoon circulation and reducing monsoon rainfall. However, GCMs have also been shown to fail to capture the relationship between surface soil moisture perturbations and cumulus convection, both because their resolution is too coarse to capture the associated mesoscale circulations, and because parameterised convection fundamentally does not respond correctly to surface soil moisture anomalies.

We carry out the first long experiment studying the effect of irrigation on pre-monsoon rainfall in India using a 4.4 km resolution, convection permitting model. We have carried out both forecast experiments and free-running simulations which allows us to see the effect of irrigation when the synoptic scale circulation is constrained and the effect when the synoptic scale can change substantially.

In the pre-monsoon, we find that irrigation increases rainfall. Intriguingly, this rainfall increase mostly does not occur on the irrigated region, but on the mountains near the irrigation. This is because irrigation, which occurs in low-lying regions, enhances the mountain-valley flows leading to enhancement of diurnal, orographically driven rainfall. Because irrigation occurs near mountains that already have some of the highest rainfall rates in the world, and which are subject to flash flooding and landslides, this has significant implications for hazards in mountainous regions during the pre-monsoon and early monsoon period.

Finally, we argue that our results are not necessarily contradictory with previous GCM studies, as the effect of irrigation likely changes with season and depends on the location and scale investigated. We suggest that further convection-permitting studies could bring to light the local impacts of irrigation on monsoon rainfall and associated risks, including drought, flooding, and landslides.