Thesis submitted to the International Institute for Geo-Information Science and Earth Observation, in partial fulfilment of the requirementsfor the Master of Science Degree in Geo-Information Science and Earth Observation, (Sustainable Agriculture).

The need for updated thematic layers describing the land cover and use is the first priority for the applications that are dealing with land exploitation and nature preservation. The opportunity to update our knowledge of the land cover and use of an area in a quick, relative inexpensive, comprehensive and extensive way, is provided by the remote sensing products. The precise filtering of the information registered on imageries and the sorting out of the surface objects is carried out with image analysis algorithms.

The accuracy of determination of the local risk potential for the use of pesticides in agricultural areas depends among other things on the available spatial information. Due to the low spatio-temporal resolution of most of the area-wide available geodata, the development of decision support systems on the farm-level generally is limited by the input data. In Germany coloured digital aerial images with a high spatio-temporal resolution (hr-images) are operational and area-wide available.

The European Union (EU) grants financial aid to farmers, producing a certain kind of useful crop. In order to administrate and to control the farmers' declarations, the EU Commission asked the member states to establish an Integrated Administration and Control System (IACS) in the year 1992. During the years of experience with the system it was found, that a high percentage of declared areas were incorrect, compared to the real situation. Therefore, the process of declaration has to be improved by the establishment of a Land Parcel Identification System, preferably based on Orthophotos.

The revised National Parks Act (1) identifies the maintenance and restoration of ecological integrity (EI) as the ‘first priority’ for park management. The approach taken is to manage for EI in the context of all aspects of the Parks Canada mandate, including EI, visitor experience and visitor education. Park managers also strive to apply concepts of ecosystem-based management for conservation of ecological values in national parks.

For agricultural management, it is important to establish GIS and maintain GIS data. It is necessary to develop the agricultural GIS in a short time and at low cost. The agricultural GIS is difficult to build up only agricultural side, and it must be linked with natural disaster mitigation system. For the same reason, data collection of ground survey is not suitable at the agricultural GIS. Remote sensing techniques maybe fit the data collection and utilization of satellite data is recommended.

Patchiness is often considered a defining quality of ecosystems in arid and semiarid regions. The spatial distribution of vegetation patches andsoil nutrients coupled with wind and water erosion as well as biotic processes are believed to have an influence on land degradation. A geostatistical measure of spatial “connectivity” is presented to directly measure the size of patches in the landscape from a raster data set. Connectivity is defined as the probability that adjacent pixels belong to the same type of patch.

In arid regions of the southwestern US, grass cover is typically a mixture of green and senescent plant material. It is important that both types of vegetation can be quantified for land management purposes and for assessing the nutritional value of grasses. Traditional ground sampling procedures are commonly used but are time consuming. Our goal was to develop an image analysis approach for separating and quantifying green and senescent grasses in the same plot using very high resolution ground photography.

This is a review of techniques for separating land cover classes for a change detection project in the Central Valley, California. The Mid Pacific Region of the U.S. Bureau of Reclamation is in the midst of the second cycle in monitoring change over major portions of the Central Valley. In agricultural areas, a considerable amount of spectrally identified change is due to cultural practices. This has led to confusion in separating change due to agricultural practice from change between categories of bare ground, grassland, forbs, and shrub land.

We investigated the capabilities of multi-temporal and multi-platform remotely sensed imagery to differentiate crop types in the 14,600 ha Upper Tiffin watershed in southeastern Michigan, a primarily agricultural area. We focused on extracting signatures for corn, soybeans, wheat, alfalfa, and grasses as the major crops for the area. Input data included Landsat 5 TM, Terra/MODIS, and ASTER imagery for different parts of the 2004 and 2005 agricultural growing season.