Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Inde → 2017 → Mapping and modeling groundnut growth and productivity in rainfed areas of Tamil Nadu

Tamil Nadu Agricultural University Coimbatore (2017)

Mapping and modeling groundnut growth and productivity in rainfed areas of Tamil Nadu

Deiveegan, M

Titre : Mapping and modeling groundnut growth and productivity in rainfed areas of Tamil Nadu

Auteur : Deiveegan, M

Université de soutenance : Tamil Nadu Agricultural University Coimbatore.

Grade : DOCTOR OF PHILOSOPHY (AGRICULTURE) 2017

Résumé partiel
A research study was conducted at Tamil Nadu Agricultural University, Coimbatore during kharif and rabi 2015 to estimate groundnut area, model growth and productivity and assess the vulnerability of groundnut to drought using remote sensing techniques. Multi temporal Sentinel 1A satellite data at VV and VH polarization with 20 m spatial resolution was acquired from May, 2015 to January, 2016 at 12 days interval and processed using MAPscape-RICE software. Continuous monitoring was done for ground truth on crop parameters in twenty monitoring sites and validation exercise was done for accuracy assessment. Input files on soil, weather and management practices were generated and crop coefficients pertaining to varieties were developed to assess growth and productivity of groundnut using DSSAT CROPGRO-Peanut model. Outputs from remote sensing and DSSAT model were assimilated to generate LAI thereby groundnut yield spatially and validated against observed yields. Being a rainfed crop, vulnerability of groundnut to drought was assessed integrating different meteorological and spectral indices viz., Standardized Precipitation Index (SPI), Normalized Difference Vegetation Index (NDVI) and Water Requirement Satisfaction Index (WRSI).Spectral dB curve of groundnut was generated using temporal multi date Sentinel 1A data. A detailed analysis of temporal signatures of groundnut showed a minimum at sowing and a peak at pod development stage and decreasing thereafter towards maturity. Groundnut crop expressed a significant temporal behaviour and large dynamic range (-11.74 to -5.31 in VV polarization and -20.04 to -13.05 in VH polarization) during its growth period. Groundnut area map was generated using maximum likelihood classifier integrating multi temporal features with a classification accuracy of 87.2 per cent and a kappa score of 0.74. The total classified groundnut area in the study districts was 88023 ha covering 17817 and 22582 ha in Salem and Namakkal districts during kharif 2015 while Villupuram and Tiruvannamalai districts accounted for 22722 and 24903 ha respectively during rabi 2015. Blockwise statistics on groundnut area during both seasons were also generated. To model growth and productivity of groundnut in DSSAT, weather and soil input files were generated using weatherman and ‘S’ build respectively besides deriving genetic coefficients for CO 6, TMV 7 and VRI 2 varieties of groundnut. Growth and development variables of groundnut were simulated using CROPGROPeanut model i.e., days to emergence (7-9 days) and anthesis (25-32 days), canopy height (63 to 70 cm), maximum LAI (1.12 to 3.07) and biomass (4176 to 9576 kg ha-1 across twenty monitoring locations spatially. The resultant pod yield was simulated to be 1796 to 3060 kg ha-1 with a harvest index of 0.28 to 0.43. On comparison of LAI between observed (2.01 to 4.05) and simulated values (1.12 to 3.07) the CROPGRO-Peanut model was found to under estimate the values with R2, RMSE and NRMSE of 0.82, 1.10 and 34 per cent. However, the model predicted the biomass of groundnut with an agreement of 89 per cent through the simulated values of 4176 to9576 kg ha-1 as against the observed biomass to 4620 to 9959 kg ha-1.

Présentation (ICRISAT)

Version intégrale (14 Mb)

Page publiée le 22 novembre 2019, mise à jour le 30 janvier 2021