RESOURCE ASSESSMENTS AND LAND DEGRADATION MONITORING WITH EARTH OBSERVATION SATELLITES

摘要

The advantages of satellite remote sensing result from its synoptic nature, comprehensive spatial information and objective, repetitive coverage. It has been illustrated how remotely sensed primary parameters, such as the spectral surface reflectance ρ, can be converted into a standardised characterisation of soil conditions and vegetation abundance. In this context, the term of thematic concepts has been introduced, by which we understand the conceptual background for identifying functional links between surface reflectance and vegetation and soil characteristics. Such concepts are primarily based on research in geosciences and ecology, and it is important to keep strong links between these disciplines and remote sensing specialists. While remote sensing has initially been used primarily for resource mapping and inventory it turns out that monitoring and predictive modelling is becoming more important and successful. However, no matter whether one intends to map surface properties or the objective is to estimate fluxes based on remote sensing data requires that the primary parameters, such as the spectral reflectance or surface temperature, are first retrieved with adequate precision. This raises immediately the issue of adequate radiometric corrections. While atmospheric effects have been a significant factor in the failure of scene models (in particular for monitoring concepts) much progress has been achieved in the past years, and we are now in the position that large parts of the radiometric preprocessing are considered a routine operation, similar to the geometric rectification. Presently, the remaining problems in retrieving surface reflectance from satellite data appear more related to the absolute radiance calibration of the sensor systems than to methodological drawbacks. This ensures that more advanced scene models, such as the spectral mixing paradigm or invertible physically-based analytical models, can be used to derive quantitative estimates and improved indicators for land resources and degradation processes. Concerning the case studies presented in this paper, a synoptic interpretation of all processed data has allowed to draw first conclusions concerning the dynamic development of selected Mediterranean vegetation communities. The fact that it has been possible to identify a degradational reduction of vegetative cover in specific locations with continued grazing pressure, implies that remote sensing can provide substantial contributions to a more conscious management of precious lands: available resources are limited, and man has already frequently crossed sensible thresholds without taking note in time. It is believed that thorough assessments of available resources, the implementation of adequate management strategies and efficient approaches to monitor and control the state of the environment are core elements on which to build efficient strategies to mitigate land degradation and desertification risks. We have shown that both inventory and mapping is required to define the current status of soil and vegetation resources and provide a baseline for monitoring, and that surveys must be repeatable and comparable, and therefore demand a standardised methodological framework. Although it is unrealistic that remote sensing will replace traditional sources of data for inventory and monitoring there is, without any doubt, an obvious role it has to play in assessing and monitoring the state of the environment. It thus forms the basis for drafting and implementing efficient land management plans which are needed to avoid land degradation under inadequate management practises.