The full-scale invasion by Russia to Ukraine in February 2022 has had a paramount impact on all spheres of life in Ukraine, including agriculture. Russia's war of aggression resulted in substantial destruction of agriculture infrastructure, reduction of crop production and available resources, redistribution of crops and high uncertainties surrounding the future of the agriculture sector in general. The full impact of the war is yet to be fully understood. In the context of land use science, substantial changes are taking place in the agricultural fields. The war is characterized by the use of heavy artillery which resulted in fields being burned, abandoned and ultimately unplanted and/or unharvested. While there exists some information on agricultural activities in free (unoccupied) territories obtained by the Ukrainian government through limited farmer surveys, no information is currently available for Russian-held territories. Earth observation data from space is the only source of synoptic, regular and objective information to understand and assess the impact of war on the national scale. The main objective of the proposed project is to advance the science of remote sensing in using infrared spectrum in combination with visible/NIR to detect and map war-induced damage to agricultural fields in Ukraine. We will provide a new cropland land-use nomenclature and develop methods to identify and map them. Specifically, the proposed nomenclature will include planted or non-planted (abandoned, fallow) fields; damaged fields and the type of damages (presence of craters, burned/non-burned); harvested or non-harvested (abandoned after planting) fields; presence of crop residue or tillage. We previously developed a deep-learning method to detect artillery craters in VHR imagery using RGB spectral bands only. We will advance detection of craters by: (i) using thermal bands (thermal anomalies) from MODIS/VIIRS, Landsat and ECOSTRESS as a proxy of potential shelling locations; (ii) incorporating SWIR spectral bands in the VHR imagery acquired by Maxar systems (e.g., WorldView-3 with eight SWIR bands); (iii) developing data fusion methods between irregular VHR Maxar’s WorldView-3 and regular moderate resolution Sentinel-2 to detect crater locations. We will develop deeplearning methods and assess the importance of SWIR bands to: (i) map planted/nonplanted, harvested/non-harvested fields; (ii) map fields with active fires; (iii) map fields with burned areas. By incorporating maps with artillery craters we will be able to characterize fields with burnings that were burned due to the shelling or typical agricultural practices. We propose to use deep-learning models, such as U-Net, in order to capture spatial context of the activities occurring in the fields as opposed to traditional pixel-based methods based on spectra indices. The proposed proposal will result in the following novelties: (i) advancing LCLUC science in characterizing, detecting and mapping damaged agricultural fields in conflicted areas; (ii) advance the use space-borne infrared spectrum for mapping the state of damaged agricultural fields; (iii) new data fusion models to detect and map artillery craters in agricultural fields using VHR and moderate spatial resolution satellite imagery.