UBC Theses and Dissertations
A study of the grid square method for estimating mean annual runoff Obedkoff, William
With the increasing importance of network planning for water resource management and inventory of supply of water there is need for new analytical methods of estimating flows from sparsely gauged regions. A new approach to estimating mean annual runoff was proposed by Solomon et al. and reported in "Water Resources Research" journal, Volume 4, October 1968. In this technique both meteorological and hydrological information are used to assess the mean annual precipitation, temperature and runoff distribution over large areas. The study area is broken up into a large number of squares and physiographic parameters are determined for each square; available meteorological data are used to derive multiple linear regression equations which relate precipitation and temperature to physiographic parameters and from these equations precipitation, temperature and evaporation are estimated for each square; runoff is obtained by subtracting evaporation from precipitation for each square and the runoff from all the squares is summed to obtain an estimate of the runoff for the entire basin; if the computed runoff disagrees with the recorded runoff, the precipitation for each square is adjusted and the procedure is repeated until the computed runoff approaches the observed runoff to the desired degree. The method has already been applied to a region in British Columbia with promising results. In the following study, use of the available basic data have been made to develop a seasonal estimate approach to the "grid square" method and in particular to consider the evaporation component and the possible incorporation of snow course data, two components which have not yet been adequately developed for use in the method under British Columbia conditions. Considering the evaporation component, it was found that apart from Turc's formula, used in the original grid square method, the Thornthwaite evapotranspiration method was the only other practical method for estimating evapotranspiration over wide areas as required by the grid square method. An attempt at an independent comparison of the two methods on an evaporation basis alone proved to be inconclusive due to the lack of adequate data but a comparison in actual computer trials of the grid square method showed that on basis of the first estimate of runoff distribution the Thornthwaite approach gave significantly better results. To incorporate the snow course data into the grid square method several approaches were taken in which an attempt at estimating on a seasonal basis the melt prior to April 1st, the date of snow surveys, was unsuccessful but showed insignificant melt which was subsequently ignored and an attempt at estimating annual precipitation at snow courses to supplement the meteorological station data was also unsuccessful. However, an attempt in which the snow course data was added to a segregated winter precipitation estimate at the meteorological stations proved to be successful and gave a small but significant improvement to the first estimate of regional precipitation and runoff distribution thus amplifying the potential use of snow course data in supplementing meteorological data for defining more clearly the regional variation of precipitation.
Item Citations and Data