Weather based turmeric yield prediction model using principal component analysis (PCA)

Abstract

Turmeric (Curcuma longa L.) is an annual herbaceous spice crop of Zingiberaceae family originated in India. India is the largest producer and consumer of this aromatic underground rhizome. Weather is one of the major limiting factors in crop production. Nowadays a large variation in crop production and yield was being observed due to climate change. This drastic change in yield prevents the farmers and policy makers to take necessary actions in the field and market. Crop weather models are statistical tools that help to represent complex relationship between crop and weather parameters and also help to predict crop yield. A field experiment was carried out on turmeric variety Kanthi at College of Agriculture, KAU, Vellanikkara during 2020 to 2021. The experiment was laid out in split plot design with four different dates of plantings (1st May, 15th May, 1st June and 15th June) as main plot treatments and four different mulches (white polythene mulch, black polythene mulch, paddy straw mulch and green leaf mulch ) as subplot treatments. There was a significant difference in yield with respect to dates of planting and different mulches applied. Yield on May 1st (24388.09 kg ha-1) and May 15th (24891.41 kg ha-1) were higher and was on par, but were significantly different from other dates of plantings. In mulching treatments, paddy straw mulch (23952.73 kg ha-1) produced superior yield. Due to the multi co-linearity, principal component analysis was utilized to find the possible linear combination of weather variables that can produce a large variance without much loss of information. With the help of principal component analysis crop yield prediction models were developed for each phenophase of turmeric.