Predicting wheat yield through artificial intelligence and crop growth simulation modelling in Punjab
Loading...
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Punjab Agricultural University
Abstract
The present study “Predicting wheat yield through artificial intelligence and crop growth simulation modelling in Punjab” was conducted at the Research Farm, Department of Climate Change and Agricultural Meteorology, Punjab Agricultural University, Ludhiana, during the rabi season of 2022-23. The field experiment was conducted with three wheat varieties (V1: WH 1105, V2: Unnat PBW 550 and V3: PBW 590), three dates of sowing (D1: 25th October, D2: 15th November and D3: 5th December) and with two microclimatic interventions (Mo: control and M1: thiourea spray @ 1000 ppm). The experiment was laid out in “split plot design” with dates of sowing and cultivars in main plots and microclimatic interventions in sub-plots. Among all the three dates of sowing, October 25th exhibited the highest average soil temperature followed by November 15th, whereas, the lowest soil temperature was recorded under December 5th. Among the micrometeorological observation, it was found that crops sown on November 15th had the highest photosynthetically active radiation (PAR) interception. Among various cultivars, Unnat PBW 550 excelled in this regard. Regrading canopy temperature, the highest average temperature was observed on December 5th, followed by October 25th, while the lowest average temperature occurred on November 15th. Among the cultivars, PBW 590 had the highest maximum average canopy temperature, followed by WH 1105, whereas Unnat PBW 550 had the lowest temperature. Under all the dates of sowing, WH 1105 exhibited the longest duration to reach physiological maturity, attributed to its extended growth period compared to the other cultivars. The GDD requirement, HTU and PTU values were also found maximum for the cultivar WH 1105 followed by Unnat PBW 550 and PBW 590. Among the biometric parameters like chlorophyll content, plant height, dry matter accumulation, leaf area index, number of tillers and yield & yield attributing characters, highest values ware recorded for November 15th as compared to October 25th and December 5th. Cultivar-wise, Unnat PBW 550 performed best for all the biometric observations than the cultivar WH 1105 and PBW 590. Among the techniques used in artificial intelligence, MLP was the best technique and ELM was considered to be as the least effective for wheat yield prediction. Using the CERES- wheat model, the maximum observed and predicted average yield was recorded under November 15th. Among the cultivars, the maximum observed
and predicted average yield was recorded for Unnat PBW 550.
Description
Keywords
Citation
Akansha (2024). Predicting wheat yield through artificial intelligence and crop growth simulation modelling in Punjab (Unpublished M.Sc. thesis). Punjab Agricultural University, Ludhiana, Punjab, India.