Validation of satellite-based cloudburst alerts: An assessment of location precision over Uttarakhand, India

Cloudburst is a transient phenomenon, where intense rainfall occurs in a localised area in a very short period, resulting in catastrophic effects like landslides and floods. The Western Himalayas region is found to be more vulnerable to cloudburst events. In this paper, the robustness of the alerts provided by the nowcasting model NETRA (Nowcasting of Extreme Orographic Rain) has been evaluated for the Uttarakhand Region. These alerts are routinely provided over Indian Space Research Organisation (ISRO) web portal www.mosdac.gov.in. It was found that the alerts generated from NETRA compare very well with actual events with district-level accuracy. The results also indicate that May month is more vulnerable to cloudburst events. From the perspective of location, the districts of Chamoli, Rudra Prayag, and Uttarkashi are found to be the most vulnerable. Among all districts, NETRA gave the best location accuracy for the district of Uttarkashi. Moreover, Apriori-based analysis is used to find the co-occurrence patterns in a cloudburst. The co-occurrence analysis brings out the spatial linkage between districts concerning cloudburst. We find out that Pauri and Uttarkashi exhibit the maximum co-occurrence probability.