Crop Identification Using Deep Learning on LUCAS Crop Cover Photos.

Massive and high-quality in situ data are essential for Earth-observation-based agricultural monitoring. However, field surveying requires considerable organizational effort and money. Using computer vision to recognize crop types on geo-tagged photos could be a game changer allowing for the provision of timely and accurate crop-specific information. This study presents the first use of the largest multi-year set of labelled close-up in situ photos systematically collected across the European Union from the Land Use Cover Area frame Survey (LUCAS). Benefiting from this unique in situ dataset, this study aims to benchmark and test computer vision models to recognize major crops on close-up photos statistically distributed spatially and through time between 2006 and 2018 in a practical agricultural policy relevant context. The methodology makes use of crop calendars from various sources to ascertain the mature stage of the crop, of an extensive paradigm for the hyper-parameterization of MobileNet from random parameter initialization, and of various techniques from information theory in order to carry out more accurate post-processing filtering on results. The work has produced a dataset of 169,460 images of mature crops for the 12 classes, out of which 15,876 were manually selected as representing a clean sample without any foreign objects or unfavorable conditions. The best-performing model achieved a macro F1 (M-F1) of 0.75 on an imbalanced test dataset of 8642 photos. Using metrics from information theory, namely the equivalence reference probability, resulted in an increase of 6%. The most unfavorable conditions for taking such images, across all crop classes, were found to be too early or late in the season. The proposed methodology shows the possibility of using minimal auxiliary data outside the images themselves in order to achieve an M-F1 of 0.82 for labelling between 12 major European crops.

Comparing land surface phenology of major European crops as derived from SAR and multispectral data of Sentinel-1 and -2.

The frequent acquisitions of fine spatial resolution imagery (10 m) offered by recent multispectral satellite missions, including Sentinel-2, can resolve single agricultural fields and thus provide crop-specific phenology metrics, a crucial information for crop monitoring. However, effective phenology retrieval may still be hampered by significant cloud cover. Synthetic aperture radar (SAR) observations are not restricted by weather conditions, and Sentinel-1 thus ensures more frequent observations of the land surface. However, these data have not been systematically exploited for phenology retrieval so far. In this study, we extracted crop-specific land surface phenology (LSP) from Sentinel-1 and Sentinel-2 of major European crops (common and durum wheat, barley, maize, oats, rape and turnip rape, sugar beet, sunflower, and dry pulses) using ground-truth information from the "Copernicus module" of the Land Use/Cover Area frame statistical Survey (LUCAS) of 2018. We consistently used a single model-fit approach to retrieve LSP metrics on temporal profiles of CR (Cross Ratio, the ratio of the backscattering coefficient VH/VV from Sentinel-1) and NDVI (Normalized Difference Vegetation Index from Sentinel-2). Our analysis revealed that LSP retrievals from Sentinel-1 are comparable to those of Sentinel-2, particularly for winter crops. The start of season (SOS) timings, as derived from Sentinel-1 and -2, are significantly correlated (average r of 0.78 for winter and 0.46 for summer crops). The correlation is lower for end of season retrievals (EOS, r of 0.62 and 0.34). Agreement between LSP derived from Sentinel-1 and -2 varies among crop types, ranging from r = 0.89 and mean absolute error MAE = 10 days (SOS of dry pulses) to r = 0.15 and MAE = 53 days (EOS of sugar beet). Observed deviations revealed that Sentinel-1 and -2 LSP retrievals can be complementary; for example for winter crops we found that SAR detected the start of the spring growth while multispectral data is sensitive to the vegetative growth before and during winter. To test if our results correspond reasonably to in-situ data, we compared average crop-specific LSP for Germany to average phenology from ground phenological observations of 2018 gathered from the German Meteorological Service (DWD). Our study demonstrated that both Sentinel-1 and -2 can provide relevant and at times complementary LSP information at field- and crop-level.

Detecting flowering phenology in oil seed rape parcels with Sentinel-1 and -2 time series.

A novel methodology is proposed to robustly map oil seed rape (OSR) flowering phenology from time series generated from the Copernicus Sentinel-1 (S1) and Sentinel-2 (S2) sensors. The time series are averaged at parcel level, initially for a set of 229 reference parcels for which multiple phenological observations on OSR flowering have been collected from April 21 to May 19, 2018. The set of OSR parcels is extended to a regional sample of 32,355 OSR parcels derived from a regional S2 classification. The study area comprises the northern Brandenburg and Mecklenburg-Vorpommern (N) and the southern Bavaria (S) regions in Germany. A method was developed to automatically compute peak flowering at parcel level from the S2 time signature of the Normalized Difference Yellow Index (NDYI) and from the local minimum in S1 VV polarized backscattering coefficients. Peak flowering was determined at a temporal accuracy of 1 to 4 days. A systematic flowering delay of 1 day was observed in the S1 detection compared to S2. Peak flowering differed by 12 days between the N and S. Considerable local variation was observed in the N-S parcel-level flowering gradient. Additional in-situ phenology observations at 70 Deutscher Wetterdienst (DWD) stations confirm the spatial and temporal consistency between S1 and S2 signatures and flowering phenology across both regions. Conditions during flowering strongly determine OSR yield, therefore, the capacity to continuously characterize spatially the timing of key flowering dates across large areas is key. To illustrate this, expected flowering dates were simulated assuming a single OSR variety with a 425 growing degree days (GDD) requirement to reach flowering. This GDD requirement was calculated based on parcel-level peak flowering dates and temperatures accumulated from 25-km gridded meteorological data. The correlation between simulated and S2 observed peak flowering dates still equaled 0.84 and 0.54 for the N and S respectively. These Sentinel-based parcel-level flowering parameters can be combined with weather data to support in-season predictions of OSR yield, area, and production. Our approach identified the unique temporal signatures of S1 and S2 associated with OSR flowering and can now be applied to monitor OSR phenology for parcels across the globe.

European Union crop map 2022: Earth observation's 10-meter dive into Europe's crop tapestry.

To provide the information needed for a detailed monitoring of crop types across the European Union (EU), we present an advanced 10-metre resolution map for the EU and Ukraine with 19 crop types for 2022, updating the 2018 version. Using Earth Observation (EO) and in-situ data from Eurostat's Land Use and Coverage Area Frame Survey (LUCAS) 2022, the methodology included 134,684 LUCAS Copernicus polygons, Sentinel-1 and Sentinel-2 satellite imagery, land surface temperature and a digital elevation model. Based on this data, two classification layers were developed using a Random Forest machine learning approach: a primary map and a gap-filling map to address cloud-covered gaps. The combined maps, covering 27 EU countries, show an overall accuracy of 79.3% for seven major land cover classes and 70.6% for all 19 crop types. The trained model was used to derive the 2022 map for Ukraine, demonstrating its robustness even in regions without labelled samples for model training.

From regional to parcel scale: A high-resolution map of cover crops across Europe combining satellite data with statistical surveys.

The reformed Common Agricultural Policy of 2023-2027 aims to promote a more sustainable and fair agricultural system in the European Union. Among the proposed measures, the incentivized adoption of cover crops to cover the soil during winter provides numerous benefits such as improved soil structure and reduced nutrient leaching and erosion. Despite this recognized importance, the availability of spatial data on cover crops is scarce. The increasing availability of field parcel declarations in the European Union has not yet filled this data gap due to its insufficient information content, limited public availability and a lack of standardization at continental scale. At present, the best information available is regionally aggregated survey data, which although indicative, hinders the development of spatially accurate studies. In this work, we propose a statistical model relating Sentinel-1 data to the existence of cover crops at the 100-m spatial resolution over the entirety of the European Union and United Kingdom and estimate its parameters using the spatially aggregated survey data. To validate the method in a spatially-explicit way, predictions were compared against farmers' registered declarations in France, where the adoption of cover crops is widespread. The results indicate a good agreement between predictions and parcel-level data. When interpreted as a binary classifier, the model yielded an Area Under the Curve (AUC) of 0.74 for the whole country. When the country was divided into five regions for the evaluation of regional biases, the AUC values were 0.77, 0.75, 0.74, 0.70, and 0.65 for the North, Center, West, East, and South regions respectively. Despite limitations such as the lack of data for validation outside France, and the non-standardized nomenclature for cover crops among Member States, this work constitutes the first effort to obtain a relevant cover crop map at a European scale for researchers and practitioners.

Harmonised LUCAS in-situ land cover and use database for field surveys from 2006 to 2018 in the European Union.

Accurately characterizing land surface changes with Earth Observation requires geo-located ground truth. In the European Union (EU), a tri-annual surveyed sample of land cover and land use has been collected since 2006 under the Land Use/Cover Area frame Survey (LUCAS). A total of 1351293 observations at 651780 unique locations for 106 variables along with 5.4 million photos were collected during five LUCAS surveys. Until now, these data have never been harmonised into one database, limiting full exploitation of the information. This paper describes the LUCAS point sampling/surveying methodology, including collection of standard variables such as land cover, environmental parameters, and full resolution landscape and point photos, and then describes the harmonisation process. The resulting harmonised database is the most comprehensive in-situ dataset on land cover and use in the EU. The database is valuable for geo-spatial and statistical analysis of land use and land cover change. Furthermore, its potential to provide multi-temporal in-situ data will be enhanced by recent computational advances such as deep learning.

Social capital and transaction costs in millet markets.

In sub-Saharan Africa, transaction costs are believed to be the most significant barrier that prevents smallholders and farmers from gaining access to markets and productive assets. In this study, we explore the impact of social capital on millet prices for three contrasted years in Senegal. Social capital is approximated using a unique data set on mobile phone communications between 9 million people allowing to simulate the business network between economic agents. Our approach is a spatial equilibrium model that integrates a diversified set of data. Local supply and demand were respectively derived from remotely sensed imagery and population density maps. The road network was used to establish market catchment areas, and transportation costs were derived from distances between markets. Results demonstrate that accounting for the social capital in the transaction costs explained 1-9% of the price variance depending on the year. The year-specific effect remains challenging to assess but could be related to a strengthening of risk aversion following a poor harvest.