Mapping rice area and yield in northeastern asia by incorporating a crop model with dense vegetation index profiles from a geostationary satellite
| Autor: | Jong-Min Yeom, Seungtaek Jeong, Ravinesh C Deo, Jonghan Ko |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: | |
| Zdroj: | GIScience & Remote Sensing, Vol 58, Iss 1, Pp 1-27 (2021) |
| Druh dokumentu: | article |
| ISSN: | 1548-1603 1943-7226 15481603 |
| DOI: | 10.1080/15481603.2020.1853352 |
| Popis: | Acquiring accurate and timely information on the spatial distribution of paddy rice fields and the corresponding yield is an important first step in meeting the regional and global food security needs. In this study, using dense vegetation index profiles and meteorological parameters from the Communication, Ocean, and Meteorological Satellite (COMS) geostationary satellite, we estimated paddy areas and applied a novel approach based on a remote sensing-integrated crop model (RSCM) to simulate spatiotemporal variations in rice yield in Northeastern Asia. Estimated seasonal vegetation profiles of plant canopy from the Geostationary Ocean Color Imager (GOCI) were constructed to classify paddy fields as well as their productivity based on a bidirectional reflectance distribution function model (BRDF) and adjusted normalized difference vegetation indices (VIs). In the case of classification, the overall accuracy for detected paddy fields was 78.8% and the spatial distribution of the paddy area was well represented for each selected county based on synthetic applications of dense-time GOCI vegetation index and MODIS water index. For most of the Northeast Asian administrative districts investigated between 2011 and 2017, simulated rice mean yields for each study site agreed with the measured rice yields, with a root-mean-square error of 0.674 t ha−1, a coefficient of determination of 0.823, a Nash-Sutcliffe efficiency of 0.524, and without significant differences (p-value = 0.235) according to a sample t-test (α = 0.05) for the entire study period. A well-calibrated RSCM, driven by GOCI images, can facilitate the development of novel approaches for the monitoring and management of crop productivity over classified paddy areas, thereby enhancing agricultural decision support systems. |
| Databáze: | Directory of Open Access Journals |
| Externí odkaz: |
|