Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #316409
Title: Coupling of phenological information and synthetically generated time-series for crop types as indicator for vegetation coverage information

Author
item MOLLER, MARKUS - Collaborator
item GERSTMANN, HENNING - Collaborator
item Gao, Feng
item FORSTER, MICHAEL - Collaborator
item THURKOW, DETLEF - Collaborator

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 3/15/2015
Publication Date: 7/22/2015
Citation: Moller, M., Gerstmann, H., Gao, F.N., Forster, M., Thurkow, D. 2015. Coupling of phenological information and synthetically generated time-series for crop types as indicator for vegetation coverage information [abstract]. 8th International Workshop on the Analysis of Multitemporal Remote Sensing Images, July 22-24, 2015, Annecy, France.

Interpretive Summary:

Technical Abstract: It is widely believed that in Germany and Europe the risk of soil erosion by water increases as a result of changes in climate. Especially, an increase of the frequency of extreme precipitation events during phenological crop phases with reduced soil cover is very likely for the near future. A monitoring of crop- and parcel-specific risk of soil erosion requires the availability of information in high spatial and temporal resolution about soil coverage by crops and crop residues. Both fractional (green) vegetation coverage (FV C) and crop residue coverage (CRC) can be estimated from spectral index profiles using linear regression models. Each of the two components requires reflectance data at specific wavelengths. One major limitation for this approach is the relatively coarse temporal resolution of the currently freely available satellite sensors that provide reflectance data of the required spectral regions at high geometric resolution. Until sensors like Sentinel 2 provide data of the required high temporal, spatial and spectral resolution, data fusion is one approach to overcome this lack of temporal resolution. During a vegetation cycle of crop types, the relative coverages differ as consequence to phenological development and phases. The vegetation coverage is relatively low on early phenological phases and increases until maximum vitality of the plants. Shortly after harvest, crop residue coverage is high but decreasing rapidly due to disintegration of the senescent plant components. Extensive phenological information can thus be used as indicator for time frames of high erodibility. They can be derived by the application of phenological models based on observations. In this study, we present an approach which couples modeled plant phenological phases with NDV I profiles derived from synthetically generated Landsat time-series of high temporal resolution based on the STARFM algorithm. The approach is applied on the spring and summer phenological phases of Winter Wheat in 2011 and visualized by a WebGIS environment. Vegetation indices show a high correlation to fractional vegetation cover, which is a major parameter to assess soil erosion risk on agricultural fields. Vegetation coverage during phases that are characterized by high NDVI values are most likely higher than those during phases of lower NDV I values. Thus erosion risk is most likely lower during these phases compared to phases of lower FV C values. Currently, we are extending this approach for the derivation if crop residue coverages. In addition, an OpenSource WebGIS is under development that will enable to determine phase-specific vegetation coverages for fields of usersupplied land use information.