Data and Methodology
In the present analysis, extreme rainfall events and flood risk are analyzed using data from well-distributed rain gauge stations during the 110-year period (1901 -
- 2010). To study the changes in the frequency of rainfall events of different intensities, the following categories of rainfall intensity based on daily rainfall are followed:
- 1. Wet days: 24-h rainfall > 0.1 mm. Alternately dry days if the day is not wet or rainfall = 0.0 mm.
- 2. Very light rain: 0.1 mm < 24-h rainfall < 2.4 mm.
- 3. Light-to-moderate rain: 2.5 mm < 24-h rainfall < 64.4 mm.
- 4. Heavy rainfall: 64.5 mm < 24-h rainfall < 124.4 mm.
- 5. Very heavy rainfall: 124.5 mm < 24-h rainfall < 244.4 mm.
- 6. Extremely heavy rainfall: 24-h rainfall > 244.5 mm.
Daily rainfall data of all the available stations during the period 1901-2010 are collected from the National Data Centre of the India Meteorological Department. From the list of all available 9294 stations, we have computed month-wise frequency of all the above six events for each of the 30 states of the country. From the monthly data, seasonal [four seasons, viz. winter (January-February), pre-monsoon (March-May), SW monsoon (June-September), and post-monsoon (October- December)] and annual time series of the frequency of all the six rainfall events for all the states are constructed. Linear trend analysis was carried out on the percentage frequency time series for each state of the six events to see the changes in the percentage frequencies of the events. In this way, we have considered all the occurrences of rainfall events throughout the country without rejecting/losing any events including significant ones.
In addition to station rainfall data, we have also used high-resolution 0.25° x
0.25° gridded daily rainfall data set (Pai et al. 2014) and the 1° x 1° gridded rainfall data of Rajeevan et al. (2008) to see the significant trends if any in the frequency of extreme rainfall events. With the gridded data, a threshold of 15 cm was considered as the extreme rainfall.