Water deficit stress limits the functioning and the performance of cultivated plants more than any other environmental factor. Pre-visual detection of how water stress varies over space and in time is thus of prime importance for agricultural crop monitoring and management. Pre-visual detection of water stress has been achieved with some success using thermal infrared (TIR) remote sensing. However, operationally available TIR image data is limited by its coarse spatial resolution, small number of spectral bands, and inadequate revisit time. Algorithms for water-deficit detection suffer from an often low accuracy of land surface temperature retrieval, unknown emissivity properties of surface material, surface heterogeneity, a lack ofsensitivity and robustness of water stress indices and a general poorunderstanding of how scale affects the uncertainty of these algorithms.Hyperspectral TIR instruments for ground and airborne use have thepotential to overcome some of these limitations. These new innovativesystems facilitate new research lines as they can acquire multiple TIRdatasets of high spatial and spectral resolution during a day and over avegetation period. Together with images acquired from drones andsatellites, scale effects on surface temperature, emissivity and stressindices can be studied. The spatio-temporal analysis of plant water stress has the potential to provide support to assessing the impact of climatic change and land management practice on the water usage of crops. The overall aim of the PLANTSENS project is to develop new techniques and services based on Earth Observation to (1) assess agricultural crop water status and to detect water-deficit stress at parcel scale and at regional scale and to (2) facilitate and fasten decision making for an adequate management of agricultural crops by providing robust and in-time analytical methods. Specific research objectives are: i) Quantification of water-deficit stress in space and time, ii) Improved surface temperature estimates and improved quantification of stomatal conductance using a hyperspectral thermal camera system, iii) Understanding scale effects of canopy temperature and spectral emissivity. The project will use multi-scale remote sensing data to estimate canopy temperature and determine water-deficit stress in agricultural crops. At the small spatial scale image data will be acquired on the ground looking at plants grown in the greenhouse. In a greenhouse water treatment experiment the methodology of detecting water-deficit stress using TIR data is tested under controlled conditions. The medium scale will be based on airborne and drone data. Airborne surveys with TIR and VNIR sensors will add complexity to the water stress detection approach in terms of atmospheric influences on the image data and land surface heterogeneity. Airborne surveys combined with extensive field campaigns will be carried out in the Attert River catchment in Luxembourg. For added flexibility, VNIRand TIR data will also be acquired using a drone system. At the large scale and for a regional assessment of water-deficit stress in agricultural crops the method will be applied to operational TIR satellite data. Consistency of water-deficit stress maps will be cross-checked with daily maps of soil moisture. The combination of standardized water stress experiments, thermal remote sensing at different levels and sound estimation of water-deficit stress will generate a better understanding of plant water interactions at different scales. The spatial analysis of plant water stress will support the assessment of the impact of climatic change and land management practice on the water usage of crops. The developed methods and derived products of the PLANTSENS project may also be used in future services specifically adapted to user needs in the agricultural sector.