Agricultural systems worldwide are impacting soil health conditions, ultimately leading to reduced productivity. In some countries, the expansion of the agricultural frontier is affecting forested areas. The improper use of agricultural inputs and inefficient water management necessitate alternative approaches to sustain food production within a more sustainable model.

 

While increasing productivity is imperative, environmental conservation must remain a priority. Ensuring the long-term health and productivity of soils is crucial for future generations. Smart farming applications based on Earth Observation (EO) have demonstrated their potential to reduce water, fertilizer, and pesticide consumption by approximately 20% while maintaining production levels. The challenge of increasing yields in existing agricultural areas, minimizing environmental impact, optimizing input use, and reducing costs are strategic indicators guiding the path toward sustainability. Gaining a deeper understanding of crop behavior and its interaction with climate and soil will enable better decision-making and promote environmentally friendly management practices.

 

 

Most EO-based crop monitoring tools rely heavily on optical satellite data. However, in regions with persistent cloud cover, the effectiveness of these applications is significantly limited. Additionally, in an effort to maintain continuous data availability, missing data is often simulated or averaged when high-quality imagery is inaccessible, leading to projection models that deviate from reality. To facilitate the adoption of more sustainable crop management practices, a reliable and continuous time series on crop phenology and health throughout the growing season is required. Furthermore, existing EO methodologies must fully leverage the diagnostic potential of satellite signals. Finally, the large data throughput of these applications demands significant computational resources, which are not readily available by default.

 

 

Against this backdrop, the DINOSAR project (Diagnostics that Integrate Optical, Infrared, and Synthetic Aperture Radar Data) aims to develop Copernicus-based algorithms to support smart farming applications that function effectively under all weather conditions, including cloud-covered regions. A better understanding of crop growth dynamics will enable farmers to adjust input applications to match actual crop needs, thereby reducing agriculture’s environmental footprint.

 

The DINOSAR project will integrate the diagnostic capabilities of optical, infrared, and Synthetic Aperture Radar (SAR) signals. This approach will drive a paradigm shift in EO-based agricultural solutions, particularly in regions with frequent cloud cover, by fully leveraging the Copernicus infrastructure. The project's initial focus will be on developing an operational multi-sensor monitoring system for sugarcane, capable of processing large data volumes. Subsequently, this methodology will be extended to real-world agricultural applications and adapted to other crops and geographies where EO-based approaches remain underexplored. The DINOSAR trains the algorithm with an extensive ground campaign to understand the radar signal behavior and model the biomass and yield of sugar can well.

 

 

Rather than treating optical and SAR data as separate signals, DINOSAR will integrate these datasets from the early stages of the processing chain—an approach that has not been previously implemented. Additionally, the project will address the operationalization of EO-based analytics by designing and testing a fully functional processing pipeline.

 

---

Carlos Mosquera, I.A. Sp. GIS, MBA.
Country Representative, ISPA.
DINOSAR Project Consortium AGROAP, ELEAF, SARVISION, UNIVERSITY OF ALICANTE, HCP International, EURONOVIA.