作者: Markus Möller , Henning Gerstmann , Feng Gao , Thorsten Christian Dahms , Michael Förster
DOI: 10.1016/J.CATENA.2016.11.016
关键词:
摘要: Monitoring of soils used for agriculture at frequent intervals is crucial to support decision making and refining soil policies especially in the context climate change. Along with rainfall erosivity, coverage by vegetation or crop residues most dynamic factor affecting erosion. Parcel-specific information can be derived satellite imagery high geometric resolution. However, their usable number mostly, due cloud cover, not representative phenological characteristics vegetated classes. To overcome temporal constraints, spatial fusion models, such as STARFM, are increasingly applied derive high-resolution time series remotely sensed biophysical parameters, based on fine spatial/coarse resolution imagery, Landsat, coarse spatial/fine MODIS. In this context, current study introduces an evaluation scheme simulated index which enables assessment performance during multiple phases. The Germany-wide available predictions phases well RapidEye parcel-specific crop-type information. results show that simulation accuracy basically controlled distance between MODIS Landsat base pairs, ability actual image properly represent phase addition, we discuss potential times corresponding (1) definition windows where potentially covered no, sparse dense (2) parameterization erosion models. database thus obtained opens up new possibilities efficient monitoring, protection hazard prevention.