A Convolutional Neural Network (CNN) model was developed to increase the resolution of satellite imagery to assess groundwater quality and monitor environmental parameters. The method achieved high performance (PSNR 32.4 dB, SSIM 0.91), enabling accurate analysis of the impact of urbanization on water resources. Based on the improved imagery and traditional parameters (pH, hardness, TDS), the Groundwater Quality Index (GWQI) was calculated for 33 districts in Egypt for 2008–2020.
The integration of satellite sensing, GIS and artificial intelligence was found to significantly improve the accuracy of soil erosion analysis and forecasting and enables the identification of degradation hotspots and optimization of protective measures, including through RUSLE models and ML algorithms. The approach proved effective in developing innovative erosion control strategies, including biodegradable materials and precision farming, for sustainable soil management.
Environmentally friendly methods (crop rotation, biosecurity, resistant varieties) were proven to be able to ensure high and stable yields, but they require an integrated approach that takes into account varietal characteristics, soil and climate conditions, and agrochemicals. In order to effectively adapt these practices, in-depth field studies are being conducted to optimize their application for specific regions and varieties.
It was confirmed that forest conservation and restoration are critical for climate goals although they are not a substitute for reducing emissions. The global carbon potential of forests outside farmland and cities is estimated at 226 Gt C, with ground-based and satellite estimates agreeing to within 12%. 61% of the potential (139 Gt C) is concentrated in existing forests, where protecting ecosystems will allow them to be restored to a mature state, and 39% (87 Gt C) is in regions with lost or fragmented forests.
Soils along the Kitchener Canal in Egypt were studied, revealing critical levels of heavy metal (Ni, Mn, Co, etc.) contamination: 79.24% of the area had very high concentrations, while 19.92% had significant concentrations. The main sources were industrial waste (textiles, paints, tannery), agricultural fertilizers, and household waste. Mapping and geostatistics (PCA, GIS) enabled the identification of risk zones, which will help implement pollution mitigation strategies, including legislative regulation of fertilizer use and waste disposal.
An environmentally friendly technology for cultivating wheat varieties, taken soil and climatic conditions, was developed. The phytosanitary status of spring wheat varieties was analyzed, the biological effectiveness of biological plant protection products in cultivation technology was assessed, the resistance of spring wheat varieties to phytopathogens and nutrient removal during the use of agrochemicals were assessed.
Corn hybrids (e.g., L101 × L105, L104 × L107) resistant to late planting were identified: their yield exceeded the one of the commercial SC-10 variety, even with reduced plant height and ear parameters.
A new biological product for crop protection based on various microorganisms, such as endophytes and mycorrhizae, was developed. The use of such biological products in agriculture can be part of a sustainable approach to farming, minimizing negative environmental impacts and ensuring crop productivity.
New remote sensing models were developed to assess and monitor grain crop cultivation. These models improve the efficiency of site management, reduce risks and optimize grain crop cultivation processes.
New models were developed to assess the risks of pesticide use in inter-row cultivation systems for leguminous crops. This helps agricultural businesses and organizations reduce the risks associated with pesticide use and make more informed decisions about crop pest and disease management.
A methodology for determining and verifying greenhouse gas emissions during the manufacture and repair of vehicle components and bodies at road infrastructure facilities was tested. Key principles for determining greenhouse gas emissions in the transport system were developed. They use and improve a three-tier methodology for assessing greenhouse gas emissions during the manufacture and repair of vehicle components and bodies at road infrastructure facilities. A review of measures aimed at reducing greenhouse gas emissions from road service facilities was done, and the effectiveness of these measures was assessed.
Non-contact methods for monitoring soil properties based on remote sensing were developed. Additional field data on the spectral reflectance (SRC) of soils in the optical and thermal ranges and their properties were collected. Models for estimating arable soil fertility parameters using remote sensing data were developed. The combined use of thermal and optical data yielded models of higher quality than using either data separately.
A scientific and technical justification for the development concept of an approximately