Parcel-based classification of agricultural crops via multitemporal Landsat imagery for monitoring habitat availability of western burrowing owls in the Imperial Valley agro-ecosystem
Agricultural production has a large impact upon the sustainability of wildlife populations. Certain species decline in the presence of intensive agriculture, while others thrive. Thus, quantifying changes in habitat within agricultural systems is paramount to understanding the underlying ecological mechanisms influencing species' range shifts and may facilitate the development of management strategies for sensitive wildlife species that depend on agricultural systems. Remotely sensed data can provide an efficient means to assess and monitor agricultural crop dynamics across large spatial extents. The objective of this study was to classify field-level agricultural crop types via a time series of Landsat imagery across the 3 620 000 ha agro-ecosystem in the Imperial Valley in California. The primary impetus was to generate data to characterize short-term changes in habitat for the western burrowing owl (Athene cunicularia), a species of concern that has an affinity for agricultural systems. The parcel-based classification method attained overall accuracies of 84% and 69% for level II (3 crop groups) and level III (31 crop types) agricultural crops, respectively. Given the large number of crop categories classified, these accuracies were quite high, especially when compared with existing crop type classifications across the region (e.g., the National Cropland Data Layer). These results suggest that the crop type classifications presented herein could potentially be used to evaluate owl demographic and space use parameters in light of shifting crop patterns across the study area. © 2010 CASI.
Canadian Journal of Remote Sensing
Parcel-based classification of agricultural crops via multitemporal Landsat imagery for monitoring habitat availability of western burrowing owls in the Imperial Valley agro-ecosystem.
Canadian Journal of Remote Sensing,
Retrieved from: https://digitalcommons.mtu.edu/michigantech-p/3214