No abstract

No abstract

No abstract

Quantifying crop production at regional scales is critical for a wide range of applications, including management and carbon cycle modeling. Remote sensing offers great potential for monitoring regional production, yet the uncertainties associated with large-scale yield estimates are rarely addressed. In this study, we estimated crop area, yield, and planting dates for 2 years of Landsat imagery in an intensive agricultural region in northwest Mexico. Knowledge of crop phenology was combined with multi-temporal imagery to estimate crop rotations throughout the region. Remotely sensed estimates of the fraction of absorbed photosynthetically active radiation (fAPAR) were then incorporated into a simple model based on crop light-use efficiency to predict yield and planting dates for wheat. Uncertainty analysis revealed that regional yield predictions varied up to 20% with the method used to determine fAPAR from satellite, but were relatively insensitive to modeled variability in crop phenology, light-use efficiency, and harvest index (the ratio of grain mass to aboveground biomass). Comparisons of satellite-based and field-based estimates indicated errors in regional wheat yields of less than 4% for both years of data. In contrast, planting date calculations exhibited uncertainties of up to 50% using a sparse, three-date sampling from satellite-based sensors. A simplified model was also developed to explore yield predictions using only one date of imagery, demonstrating high accuracies depending on the date of image acquisition. The spatial and temporal distributions of crop production derived here offer valuable information for agricultural management and biogeochemical modeling efforts, provided that their uncertainties are well understood.

No abstract