The present study was performed in order to compare different methods for calculating degree-days. Four methods of thermal sum, often used in calculation of degree-days, were compared: Ometto, Snyder, Dufault and Arnold residual methods. These methods were applied to corn and beans crops. The simulations have been done considering the actual temperatures on Viçosa, MG, Brazil, in 2011, and simulated temperatures (random additions between 0 and 5ºC for each day). The responses of the four methods on calculating the degree-days were similar for the observed temperatures in both studied crops. However, for the higher temperatures (simulated), the methods did not show similar responses. Fixing the degree days in 1,600 ºC day for corn crop and 1,300 ºC day for beans, the studied methods showed different reduction on the crop cycle. The largest reductions were obtained by Arnold and Snyder methods which have reduced equally the cycles by 28 and 17 days for corn and beans, respectively. The Ometto method resulted in the lowest reductions: 12 (corn) and 11 (beans) days, because it uses the upper and lower baseline temperatures, and it considers a greater penalty in days when the temperature exceeds the maximum baseline temperature.
degree-day; temperature; maize; beans
