Pre-harvest forecasting models for potato (Solanum tuberosum) and mustard (Brassica juncea) yield in central plain zones of Uttar Pradesh

The potato (Solanum tuberosum L.) and mustard (Brassicajuncea L.) yield predicted for the year 2013–14 using step-wise regression model to meet the objective of developing appropriate yield model of different districts. The yield data were collected from crop report published by Directorate of Agriculture, Utter Pradesh. Last 19 year's crop data of fourteen crop growing districts, viz. Agra, Auriya, Mathura, Etah, Firojabad, Farrukhabad, Hathrus, Kanpur Dehat, Kanpur Nagar, Hardoi, Mainpuri, Unnao, Etawah and Kannauj falls under central plain zones of Uttar Pradesh were used to find out relationship with weather parameters such as maximum (Tmax.) and minimum temperature (Tmin.), morning relative humidity (RH1) and evening relative humidity (RH2) and rainfall (RF) were used in regression models as input in the same district or nearby meteorological observatory. Models were validated on the basis of two years (201112 and 2012–13) data. The results revealed that models explained 0.5 to 54% variation for potato yield and zero to 68% for mustard yield in different districts. The root mean square error (RMSE) ranged from 2081.8 kg (Firojabad) to 7177.8 kg (Kanpur Dehat) for potato and from 112.2 kg (Kanpur) to 219.9 kg (Mathura) for mustard crop. The models are validated with ±10% error in all the fourteen districts. Hence, these models can be used for predicting mustard and potato yields of different districts of central Uttar Pradesh.