Empirical model for turmeric (Curcuma domestica Val.) yield prediction

Abstract

The mean weather variables namely maximum (TMAX) and minimum (TMIN) temperature, maximum (RHMAX) and minimum (RHMIN) relative humidity, rainfall (RAINF), rainy days (RAIND), evaporation (EVPN), wind speed at 8 feet height (WINDS), sunshine hours per day (SUNSH) and total solar radiation (RADN) of different months of turmeric crop season for twenty years (1979-80 to 1998-99) were correlated with yield (dry) in Coimbatore District, Tamil Nadu. Out of the ten variables studied, three variables (TMAX, RHMAX and RADN) did not show significant relationship with yield while the remaining seven variables had significant correlation with yield. These significant variables were further subjected to stepwise regression analysis and those variables, which could be related to yield variability reasonably (EVPN, RHMIN, RAINF, TMIN) were used in the final regression model. This model was used to estimate turmeric yield in Coimbatore District and it was found that the estimated and observed yields did not differ. This model (Y= -11675.5119 - 591.0617 EVPN3 + 810.3569 TMIN, + 12.1481 RAINF2 + 91.7499 RHMIN,) can be used to predict the yield of turmeric and total production in Coimbatore District, Tamil Nadu, India.