Evaluation of four degree-day estimation methods in eight Colombian coffee-growing areas

Abstract

Methods for estimating the accumulation of degree-days based on maximum and minimum temperatures are commonly used to determine relationships or to adjust phenological models based on the “physiological time”. Degree-days are obtained indirectly by these methods, given that information is not generally available on hourly or shorter time scales, due to the type of equipment used to record data or a data loss in historical time series. To compare the performance of such methods, degree-days were estimated with four indirect techniques in eight Colombian locations during one year. Each indirect technique was evaluated in comparison to the numerical integration method by the trapezoidal rule (reference method) using temperatures recorded every five minutes. Based on the percent bias, the methods proposed by Arnold, Ometto and Snyder tend to overestimate thermal time, whereas the Villa Nova method underestimates this time. The use of the different methods depends on the information available for the estimation of the study variable.

Keywords

• thermal time
• temperature
• numerical integration
• linear regression
• error