Agricultural landscapes are changing their appearance with the increasing use of man-made plastic materials in covered agriculture (plasticulture) all around the world. As these affect the landscape visually, increase pollution and decrease local biodiversity, better monitoring and planning of their uses and areas are needed. Using a field spectrometer we studied the spectral properties of a sample of polyethylene sheets and various nets used in Israel. We detected three major absorption features around 1218 nm, 1732 nm and 2313 nm. These were not affected by settling dust, whitewashing or by the underlying surface, but were not apparent in black coloured polyethylene sheets or nets. A hyperspectral AISA-ES image with a spatial resolution of 1 m achieved a detection accuracy of above 90% for bright sheets and nets but of only 70% for the black nets. The best spectral feature for plastic mapping was found to be that around 1732 nm as it does not coincide with spectral features of other minerals, soils, vegetation or atmospheric attenuation. As most of the greenhouses patches in Israel are smaller than 3200 m2, the optimal spatial resolution of a sensor for mapping them should be equal or better than 8-16 m. As a result of their low spectral and spatial resolution, Landsat images proved inadequate for mapping greenhouses, and strengthen the need of hyperspectral technology for that end.
Department of Geography and the Human Environment, Tel Aviv University, Ramat Aviv, aviv 69978, Israel 2:
Surveys Unit, Open Landscape Institute, The Society for the Protection of Nature in Israel, aviv, Israel 3:
Bar-Kal System Engineering Ltd., Israel