| Autor: |
Jarrod O. Miller, Pinki Mondal, Manan Sarupria |
| Jazyk: |
angličtina |
| Rok vydání: |
2024 |
| Předmět: |
|
| Zdroj: |
Crop and Environment, Vol 3, Iss 1, Pp 33-42 (2024) |
| Druh dokumentu: |
article |
| ISSN: |
2773-126X |
| DOI: |
10.1016/j.crope.2023.11.001 |
| Popis: |
The use of sensors for variable rate nitrogen (VRN) applications is transitioning from equipment-based to drone and satellite technologies. However, regional algorithms, initially designed for proximal active sensors, require evaluation for compatibility with remotely sensed reflectance and N-rate predictions. This study observed normalized difference vegetation index (NDVI) data from six small grain and two corn fields over three years. We employed three platforms: tractor-mounted active sensors (T-NDVI), passive multispectral drone (D-NDVI), and satellite (S-NDVI) sensors. Averaged NDVI values were extracted from the as-applied equipment polygons. Correlations between NDVI values from the three platforms were positive and strong, with D-NDVI consistently recording the highest values, particularly in areas with lower plant biomass. This was attributed to D-NDVI's lower soil reflectance and its ability to measure the entire biomass within equipment polygons. For small grains, sensors spaced on equipment booms might not capture accurate biomass in poor-growing and low NDVI regions. Regarding VRN, S-NDVI and D-NDVI occasionally aligned with T-NDVI recommendations but often suggested half the active sensor rate. Final yields showed some correlation with landscape variables, irrespective of N application. This finding suggests the potential use of drone or satellite imagery to provide multiple NDVI maps before application, incorporating expected landscape responses and thereby enhancing VRN effectiveness. |
| Databáze: |
Directory of Open Access Journals |
| Externí odkaz: |
|
