Fisher's regression integral versus regression function of selected weather factors in crop-weather analysis

Abstract

The paper deals with the relative performance of the two statistical methods, namely, (i) Fisher's regression integral' which brings out the slow continuous change in the response of crop to the weather pattern experienced by the cultivated soil and crop and (ii) regression function in which 'weather pattern' is subjected to continuous screening to yield a few well defined weather periods of significance to the soil and crop. In the case of wheat crop at Jalgaon and Niphad the regression function has given better multiple correlation coefficient than the regression integral. This may be due to the differential response of some of the adjacent phytophases of crop and same the changing soil characteristics to the weather factors. By and large, this inference was found to be true as seen from  the physiological and pedological considerations.