University of Twente Student Theses
Biophysical parameters retrieval for pearl millet using semi-empirical model
Ramathilagam, Arun Balaji (2020) Biophysical parameters retrieval for pearl millet using semi-empirical model.
Full text not available from this repository.
| Full Text Status: | Access to this publication is restricted |
| Abstract: | Biophysical parameters retrieval is one of the important applications of remote sensing in the field of crop monitoring. Different biophysical parameters like crop biomass, height and LAI can be estimated using Water Cloud Model, a semi-empirical modelling technique. However, this model is based on assumptions that the water content is uniformly distributed throughout the canopy, which is wrong in the case of pearl millet. In India, pearl millet is usually broadcasted and so plant to plant or row to row spacing is not seen. As a result of this, the plant density varies within the fields, and due to the crop structure, the water content is not uniformly distributed within the crop canopy. Considering these limitations due to vertical and horizontal variation in water content, a modified Water Cloud Model is proposed in this study, to predict different crop biophysical parameters. The crop variables like cob, stem and vegetation biomass, LAI, and stem height were estimated using two different model inversion techniques – LUT search technique and Random Forest Regression in this study. . By LUT search technique, the RMSE of the predicted cob biomass varied from 0.337 kg/m2 (VH) to 0.514 kg/m2 (Ps), he RMSE of the predicted vegetation biomass varied from 1.226 kg/m2 (HV) 1.297 kg/m2 the RMSE of the stem biomass predicted using PS and PD were 1.109 kg/m2 and 1.105 kg/m2 respectively The RMSEs of the predicted stem height using PS, PD and PV were 0.537 m, 0.514 m and 0.567 m respectively. The LAI was predicted using PS, PD and P (VV)(, V with the RMSE 3.091, 2.436 and 1.583, respectively. Using RF regression technique, RMSE of the predicted cob biomass varied from 0.317 kg/m2 (PV) to 0.437 kg/m2 (PS), the RMSEs of the predicted vegetation biomass varied from 1.089 kg/m2 (HH+VV) to 1.308 kg/m2 (HV) and the RMSE of the predicted stem biomass varied from 0.799 kg/m2 (PD) to 0.974 kg/m2 (PS). While the stem height was predicted with the RMSE varying from 0.434 m (PD) to 0.502 m (PV), LAI was predicted with the RMSE varying from 2.315 (PS) to 1.457 (PV). |
| 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/98660 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page