Simple models to estimate soybean and corn percent ground cover with vegetation indices from MODIS
Remote sensing images are a good source of crop and soil information, which can be used to derive agronomic information for field management and yield prediction. Soybean (Glycine max (L.) Merrill) and corn (Zea mays L.) are the most important crops in Argentina taking into account the economic yiel...
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | article |
| Lenguaje: | Inglés |
| Publicado: |
2022
|
| Materias: | |
| Acceso en línea: | http://hdl.handle.net/11086/24477 http://www.aet.org.es/?q=revista39-8 |
| Aporte de: |
| Sumario: | Remote sensing images are a good source of crop and soil information, which can be used to derive agronomic information for field management and yield prediction. Soybean (Glycine max (L.) Merrill) and corn (Zea mays L.) are the most important crops in Argentina taking into account the economic yield obtained by farmers and the sown area. In this work, simple mathematical models (linear, second order polynomial and exponential), with different vegetation indices (VI) derived of Moderateresolution Imaging Spectroradiometer (MODIS) images as inputs, were evaluated. The models were applied to estimate soybean and corn percent ground cover (fCover) over the growing season. The VI employed were the normalized difference vegetation index (NDVI), soil adjusted vegetation index (SAVI), a modified SAVI (MSAVI), simple ratio (SR) and perpendicular vegetation index (PVI). The performances of the models (linear, polynomial and exponential) were very good and their results were equivalent. Although all models could successfully estimate fCover, results showed that, excepted with SR input, a linear model can predict ground coverage with R2 values greater than 0.86, when both crops are considered. When models are applied to soybean and corn separately, linear model with SAVI index has the best performance. |
|---|