Crop classification and crop water need estimation of Piave river basin by using MIVIS, Landsat-TM/ETM+ and ground-climatological data

Abstract

In this work a classification of the main irrigated crops of the Piave river basin and an estimation of crop water requirements during the growing season are presented. The work is divided into two parts. The first includes recognition, mapping and quantification of the main irrigated crops for thematic map production and a database creation. MIVIS hyperspectral airborne data, Landsat-TM/ETM+ multispectral satellite data and ground truth data were used for crop classification. A specific method of knowledge-based image classification was designed and used. The proposed method was compared with other per point conventional classification methods. In the second part the crop water need estimation is discussed. Ground-climatological data of the study area ground-climatological stations were used. The water balance equation parameters were estimated on a ten-days basis. A spatial interpolation method was used to propagate these parameters at pixel spatial resolution to study area. Soil water deficit map for irrigation was produced and a flow rate estimation was performed.