Emploi de la télédétection pour l'étude de l'humidité des sols

Teledetection for soil moisture surveys

Président du Groupe international « Pédagogie et Télédétection »

Abstract

Soil moisture is an important item of information for the agronomist in that it determines the plants' water supply. Determination of surface moisture, however, is insufficient for giving advice to a farmer; one has to reason in terms of volume. In addition, one must determine the dynamics of the moisture content of the considered volume in order to establish whether the plant will find water when it needs it. Teledetection provides data on soil moisture and associated factors. Some of the latter can be detected directly. Micro-waves are required to obtain moisture data for the top 15 cm of soil. Indications are also obtainable by infra-red heat measurement. The following can be ascertained by visible and close infra-red range measurement : 1. organic matter content (Fig. 1), 2. soil roughness (Fig. 1) and 3. moisture (Fig. 2). The 950 nm and 1150 nm bands should supply interesting results. Other factors can be detected by indirect methods. Factors connected with moisture for example can be determined from morphology, soil surface and vegetation data. To understand the "soil moisture" dynamic system, one has to consider the following: a) Soil water supply b) Soil water losses c) Water retention by the soil Having carefully studied the above, one then considerations what teledetection could contribute for each. Soil stoniness, roughness and other factors also have to be taken into account. Soil moisture can be interpreted in four stages related to the following : a) Soil water dynamics b) Estimated soil water storage c) Internal soil water circulation d) Soil water balance A partial model is constructed at each stage, the elements of which are questions which can almost invariably be answered by interpretation of relevant pictures and photographs. The models are run one after the other in order to successively integrate the static and dynamic aspects (models 1 and 2 respectively), then the spatial and temporal aspects (model 3 and water balance respectively). Using simple models, various soil moisture states can be assessed in both time and space. To run such models, one requires the full set of teledetection data and a certain number of ground-based measurement data.