A simple method for supplementing long-term weather records in a maize-modelling exercise

Open Access
Publisher Website
Google Scholar
Abstract
Missing weather data is a very real problem when using daily-incrementing crop growth simulation models. This is especially true in the case of daily air temperature and solar radiation data. The aim of this study was to compare two simple methods of supplementing the input weather record, and to determine whether the estimates could be used with confidence in the case of the CERES-Maize model. Wet and dry day monthly means of daily maximum and minimum temperature and sunshine duration were calculated for four long-term weather stations in Natal. Ordinary monthly means of these elements were also computed. The two sets of means, which were calculated from the most recent 5 years of the weather record, were appended program-matically to daily rainfall data for the preceding 10 years of the weather record, in each case replacing totally the observed daily temperature and sunshine duration values. The weather files thus created were then used as inputs to the CERES-Maize plant-growth model. Yield, flowering date and maturity date estimates so gained were compared with similar estimates over the same period using real temperature and sunshine data. Yield estimates based on supplemented data compared excellently with those using the full weather record, with mean percentage differences for the four sites ranging from 1.23% to 9.34%. Yield estimates based on wet and dry day means for the weather record were slightly better than those derived using ordinary monthly means. This was also true in the case of the flowering and maturity date estimates.

This publication has 4 references indexed in Scilit: