Quantitative precipitation forecasting for hydrological prediction over catchment areas with inadequate agrological data by the method of atmospheric water balance

Abstract

In this paper attempt has been made to estimate rate of precipitation per unit area in and around northeast India containing the catchment of river Bhagirathi by the method of tmospheric water balance as developed by Palmen.

In order to overcome the inadequacy of upper air temperature, wind and humidity data, due to the absence of a waves radiosonde / rawin station in the catchment and the paucity of a close network of such stations in the neighbourhood, radiosonde /rawin data of remote stations like Allahabad, Calcutta, Gauhati, Delhi, Nagpur, Visakhapatnam and port Blair have been made use of for working out the specific humidities at isobaric levels by drawing isopleths of specific humidities and interpolating the values at intermediate points. The isopleths themselves have been drawn with due regard to the airmass prevailing over the area during the rainstorm period. Similarly, isopleths of U and V components of the observed upper winds prevailing over the region have been drawn with reference to the synoptic situation. Upper wind data of pilot balloon stations have also been utilised in addition to those of the radiosonde/  rawin stations for the drawing of isopleths of wind. The divergence of wind has then been calculated by using finite difference method.

The rates of precipitation at different stations and consequently the isohyets of total amounts of precipitation for 9 hours have been compared with those of the observed rainfall at 0830 and 1730 hours of the same day, for individual days of the rainstorm period. The estimated precipitation for 9-hour has also been compared with the observed precipitation at 0830 IST of the next day as the calculated isohyetal patterm for 24 hours was identical with that of the 9 hours except in the magnitudes of the station precipitation. The mean observed storm precipitation has also been compared with the mean calculated amounts of precipitation for the same period as also the precipitation for the period of the storm and 1-day prior to the storm. Storm efficiencies have been calculated between the observed rainfall and the estimated amounts of rainfall. Daily and the mean vertical distribution of the horizontal transport of moisture has been plotted for the storm period.