I. Development and Validation of Leaf Wetness Duration Model. II. Leaf Wetness Duration Based Sheath Blight Prediction in Burdhaman District, West Bengal.

Leaf Wetness Duration (LWD) has been considered as one of the key parameters in agrometeorology for disease prediction, due to its power of stimulating and controlling many host-pathogens interactions causing several plant diseases. Mainly two types of mathematical models have been proposed for LWD estimation- Physical and Empirical models. Physical models provide more reliable estimation than that of empirical models and are portable in nature. So, in this study, one of the most accurate physical models, Penmann- monteith approach has been used for LWD estimation. Many LWD based disease prediction models have been proposed for different crops. Pessl model for rice sheath blight (Rhizoctonia solani) has been considered here for validating its applicability for Burdhaman district of West Bengal, India. Here, the main challenge in estimating LWD is the lack of hourly net radiation and solar radiation data which are the key inputs of physical models. Spokas’s method has been used for hourly solar radiation calculation from temperature and precipitation. Due to unavailability of solar radiation and LWD sensor data from the above mentioned study area, SAS (Strawberry Advisory System) from Florida, USA has been referred to validate LWD estimates from calculated solar radiation by Spokas’s model. Model estimation of LWD shows a good agreement on sensor observations from SAS with R-square = 0.83. It shows a negative bias or underestimation of wetness duration with a ME of -0.88h and MAE of 0.97h. These error estimates proves its potential to be extended for other locations with a good confidence. Validation of LWD based Pessl model for rice sheath blight, has been carried out using remote sensing based disease index proposed by Qin, 2005. Comparative analysis shows a good correlation. Hence we propose to use this model in real time to provide sheath blight disease severity alerts.