A comparison of methods to determine the onset of the growing season in northern Nigeria

Open Access
Publisher Website
Google Scholar
Abstract
In the sudan savanna of northern Nigeria, with its semi‐arid climate, the ability to determine effectively or predict the start of actual productive rains cannot be overemphasized. Several methods exist for calculating the date of onset of the rains that may be taken as the start of the growing season. Five methods currently in use, which are relatively easy to apply on a large scale, were selected for comparison. One is a traditional technique (Ramadan method), two use accumulated rainfall totals (Walter's and Sivakumar's methods), and two use rainfall–evapotranspiration relationships (Kowal's and Benoit's methods). For the period 1961–91, the traditional technique performed most poorly. Walter's method gave quite early onsets and Sivakumar's method gave very late onsets, thereby seriously shortening the growing season. Kowal's and Benoit's determinations fell most often in between the results of Walter's and Sivakumar's methods in their performance. However, although generally to a lesser extent than the other methods, they are still significantly affected by false starts. To avoid incorrectly predicting the growing season's onset as far as possible, but to prevent an unacceptable shortening of the growing season, a combination of Kowal's and Sivakumar's criteria was used to develop an improved technique. This proved to work well for determining the onset date in the study area. However, because some false starts remain when using average onset dates, it is suggested that an operational advisory team should be constituted by the government. This team would be responsible for calculating onset dates in any year, on‐line for the ongoing season, in a participatory approach with farmers, and for disseminating such dates to the farmers. This could be done for any place for which the appropriate data can be made available. In the future, improved climate prediction skill may replace the classical probabilistic approaches presently suffering from increasing rainfall variabilities. Copyright © 2002 Royal Meteorological Society.

This publication has 11 references indexed in Scilit: