University of Twente Student Theses

Login

Enhancing a remote sensing based crop growth model to include genetic parameters that capture dynamic site specific management aspects: A Case Study of Rice Production in the Mekong River Delta, Vietnam

Kabiri, Stella (2009) Enhancing a remote sensing based crop growth model to include genetic parameters that capture dynamic site specific management aspects: A Case Study of Rice Production in the Mekong River Delta, Vietnam.

[img] PDF
6MB
Abstract:The objective of this study is to introduce into a remote sensing based crop growth simulation model, genetic parameters that reflect impacts of site specific management aspects to improve yield prediction. It seeks to identify cultivar characteristics of rice genotypes highly influenced by the environment and can be detected by remote sensing, based on model analysis. A cross-locational survey was carried out in ten districts of the Mekong River Delta, a major rice growing region of Vietnam. The ORYZA2000 crop growth model for rice was calibrated with site specific management aspects and observed Leaf Area Index (LAI) and Specific Leaf Nitrogen (SLN) fraction in leaves obtained from red (R) and near-infrared (NIR) regions, represented in MODIS’ two 250-m spectral bands. model was simulated in two categories; initial and calibrated. Two most dominant varieties were selected, IR50404 and Jasmine 85. The enhanced crop growth simulation model explained almost 70% of the variation of the predicted yield for all the sites, 80% for Jasmine 85 and 88% of the variance of the predicted yield. The initial model poorly estimated the effect of G x E interactions on yield of both varieties in all the test environments (neutral, saline, acid and acid saline soils). The calibrated model improved the prediction in all the test environments except Jasmine 85 was over estimated in the neutral soils. Sensitivity analysis of the enhanced crop model was run by a simulation strategy based on environmental factors affecting the dynamics of photosynthesis affected by LAI and SLN. The potential yield increase was 14% for the site specific model, 0.6% for IR50404 and 2% for Jasmine 85. Sensitivity analysis also showed that the highest potential yield increase was in the acid saline soils for both SLN and LAI meaning that, these varieties were deficient in these traits for this environment. The output of the simulation shows that it is possible to include dynamic site specific management aspects and genetic parameters from remote sensing based techniques in a crop growth simulation model and satisfactorily predict yield in rice genotypes. It also shows that LAI and leaf N content detected from remote sensing based techniques are traits that can be used in elevation of yield potential in breeding programmes and reduce the duration of costly multilocational trials during cultivar improvement.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:https://purl.utwente.nl/essays/93052
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page