Hyperspectral images from newly launched (ASI-PRISMA and DLR-EnMAP) and future satellite (ESA-CHIME) are an opportunity, thanks to the high spectral resolution and full range continuity, to improve the retrieval of information about the crop parameters and status.The high dimensionality of hyperspectral data and the non-linear relationship between the crop biophysical parameters and their spectral signature make quantitative estimation of crop characteristics challenging, to address these problems we tested different configurations of neural networks (fully connected and convolutional).We tested the different architectures on two training dataset, one consists in ground data collected in three experiments, in different locations and seasons, the second one (hybrid) is composed by synthetic data generated using a radiative transfer model (PROSAIL-PRO).Preliminary results for LAI, CCC and CNC retrieval are encouraging in particular when ground data are exploited demonstrating of the potentiality of NN to fully exploit the information density of the hyperspectral data.

An algorithm to identify and monitor tillage practices, using Copernicus Sentinel-1 (S-1) and Sentinel-2 (S-2) data, is presented. The technique operates on agricultural fields that are either bare or sparsely vegetated. These fields are first segmented using the Normalized Difference Vegetation Index (NDVI), obtained from S-2, or the S-1 VH/VV ratio in overcast conditions. Then, a change detection approach is applied both to S-1 cross-polarized backscatter and copolarized interferometric coherence. To decouple the impact of tillage from that of moisture change on radar measurements, a two-scale strategy is used. The premise is that whereas soil moisture is primarily influenced by precipitation events happening at the medium (1.0-10 km) scale, tillage changes occur at the local, i.e., field (~0.1 km) scale. The algorithm was assessed against a multi-year ground data set collected at three sites. It includes conventional tillage change and no-tilled events. Results achieve an overall accuracy of 81%.

Recent studies have shown that Convolutional Neural Networks (CNNs) achieve impressive results in crop segmentation of Satellite Image Time-Series (SITS). However, the emergence of transformer networks in various vision tasks raises the question of whether they can outperform CNNs in crop segmentation of SITS. This paper presents a revised version of the Transformer-based Swin UNETR model adapted specifically for crop segmentation of SITS. The proposed model demonstrates significant advancements, achieving a validation accuracy of 96.14% and a test accuracy of 95.26% on the Munich dataset, surpassing the previous best results of 93.55% for validation and 92.94% for the test. Additionally, the model’s performance on the Lombardia dataset is comparable to UNet3D and superior to FPN and DeepLabV3. Experiments of this study indicate that the model will likely achieve comparable or superior accuracy to CNNs while requiring significantly less training time. These findings highlight the potential of transformer-based architectures for crop segmentation in SITS, opening new avenues for remote sensing applications.

This study investigates the efficacy of incoherent and coherent SAR descriptors for filling spatial and temporal gaps in optical-driven Leaf Area Index (LAI) time series. Within this context, an artificial intelligence (AI) algorithm based on Multi-Output Gaussian Process (MOGP) [1], [2] demonstrated its effectiveness in handling the different information derived from SAR signatures in a unified corpus. The study utilizes sequences of Sentinel-2 imagery to derive Leaf Area Index (LAI) maps, while Sentinel-1 observations over the same area are utilized to obtain SAR backscatter coefficients and interferometric coherence data. This comprehensive dataset is then employed as input for training the MOGP model. Experimental tests demonstrate the usefulness of the MOGP model in obtaining accurate LAI time series even during very cloudy periods.

The objective of this paper was to assess the use of multi-frequency SAR data for the mapping and monitoring of the spatial and temporal variability of land surface parameters and agricultural practices. In particular, the focus was on the retrieval of surface soil moisture (SSM) and vegetation water content (VWC) and on the classification and monitoring of irrigation extent and tillage practices at high resolution. The paper illustrates the data basis collected over three European sites, namely Apulian Tavoliere (Southern Italy), Jolanda di Savoia (Northern Italy), and Castilla y Leon (Spain), and the main results.

Earth Observation derived land cover products can provide an advance insight into coastal ecosystem monitoring to capture the fragmentation and changes of land surface. The work addresses the testing of consolidated classification algorithms on hyperspectral data in synergy with SAR and field data for the definition of biological and geophysical characteristics of surface coverage components to produce a detailed Dynamic Land Cover map. We obtain reliable land cover and land use information identifying different types of vegetation and soils condition and indicators of biogeophysical surface processes in the testing costal area of the Gulf of Oristano (Italy). The results offer further classification of specific land cover targets representing a contribution for the production of land cover products at different levels of thematic resolution to improve downstream application services and to support land policies and management.

PRISMA (PRecursore IperSpettrale della Missione Applicativa) is a demonstrative spaceborne mission, fully deployed by the Italian Space Agency (ASI). To support the calibration/validation activities of the PRISMA hyperspectral mission, ASI and the National Research Council (CNR) started in 2019 the PRISCAV project (Scientific CAL/VAL of PRISMA mission). The main objective of PRISCAV is the comprehensive characterization of the performances of the PRISMA payload in orbit in different operational scenarios and the verification of the durability in time of the performances. To this end, PRISCAV created a network of 12 instrumented sites showing different land-use and surface settings (Snow; Sea; Inland and Coastal Water; Forest and Cropland) to obtain independent and traceable in-situ and airborne Fiducial Reference Measurements (FRM) simultaneous to PRISMA acquisitions in order to assess the required performance of sensor, data products, and processors at the different levels (i.e. Top-of-Atmosphere Level 1 Radiances and Bottom-of-Atmosphere Level 2 Reflectance standard products). Moreover, on some of these sites, simultaneous PRISMA and airborne AVIRISNG acquisitions were made coupling remote sensing with in-situ observations to support new mission development and in particular the Copernicus Hyperspectral Imaging Mission for the Environment (CHIME). Recent updates on CAL/VAL activities and on AVIRSNG campaigns are presented in this contribution.

The study aims to consolidate and validate a suite of Earth Observation algorithms of interest for applications in agriculture. The algorithms are at different levels of maturity. Still, they share the objective of contributing to sustainable water management and food security. They deal with monitoring the soil moisture, the vegetation water content, the extent of irrigated areas and the changes in the surface roughness of agricultural fields. The paper introduces the data sets, the algorithms and discusses some examples of initial results.

The goal of this work is to investigate the potential of PRecursore IperSpettrale della Missione Applicativa (PRISMA) hyperspectral data to predict the concentration of four macronutrients (K, P, N, S) and four micronutrients (Ca, Fe, Mg, Zn) in final wheat production. All investigated nutrients are essential to improving human nutrition. The initial findings indicate accurate predictions for Zn, P, Mg, S, K, Ca and Fe (R2 ranging from 0.57 to 0.74). N was less accurately estimated (R2 of 0.49). We conclude that the foliar chemical properties and temporal dynamics as detected by hyperspectral data translate successfully to the target micro- and macronutrients composition of the wheat production.

Lodging in wheat is one of the main constraints limiting yield and grain quality. Accurate information about crop lodging susceptibility during the growing season is critical for improving yield estimates and for targeting the expenditure on lodging control. In this context, this study aims to estimate safety factor against root lodging ($SF_{A}$) as a measure of lodging susceptibility by exploiting Sentinel-1 data using Extreme Gradient Boosting Regression. Through extensive field experiments during a crop season, several crop variables were collected from several plots in multiple visits, and the corresponding metrics were extracted from the Sentinel-1 images. Our results show that the field measured$SF_{A}$correlated well with the field lodging and the cross-validated regression model could estimate$SF_{A}$with an$R_{CV}^{2}=0.73$and$RMSE_{CV}=0.59$. Thus, the$SF_{A}$measure constitutes a state-of-the-art approach in the remote sensing community for the assessment of root lodging susceptibility.

Non-photosynthetic vegetation (NPV) in croplands is receiving a growing interest for its relevance in the field of sustainable agriculture. Previous studies have demonstrated the suitability of hyperspectral remote sensing in detection and classification of NPV. This study is an early assessment of the PRISMA mission capability to provide quantitative estimates of NPV, through exploiting PRISMA imagery and reference field data collected on summer 2020. We investigate the cellulose absorption region (2.0-2.2 µm) applying a technique of feature reduction (Exponential Gaussian Optimization - EGO) able to retrieve spectral parameters which can be used for qualitative and quantitative assessment of NPV in croplands. In particular, a significant non-linear relationship has been found between EGO-derived band depth and dry matter abundance (g/m2). The results are very encouraging toward a robust and quantitative estimation of NPV from space and worthy of further investigations.

In this work we present a procedure for building a dataset of fire reference perimeters over the African continent from Sentinel-2 (S2) time series. The strategy relies on a sampling scheme designed on the characteristics of the S2 tiling system to provide units suitable for statistical sampling of validation units. The S2 archive is searched to extract cloud free images for building time series that are classified with a Random Forest algorithm to provide fire reference perimeters. A test S2 tile over Tropical savanna (35 LMD) is used to assess the accuracy of S2 reference perimeters by comparison with polygons of burned areas extracted from high resolution Planetscope data.

The Copernicus Hyperspectral Imaging Mission for the Environment (CHIME) is in preparation to carry a unique visible to shortwave infrared spectrometer. CHIME will globally provide routine hyperspectral observations to support new and enhanced services for, among others, sustainable agricultural and biodiversity management. The mission shall provide Level 1B, 1C and 2A products, as well a set of downstream products related to the different environmental applications, such as the quantification of vegetation traits. In this context, this work presents the first hybrid retrieval models for the operational delivery of vegetation properties products. Within ESA's CHIME end-to-end (E2E) simulator study, 13 leaf and canopy trait models were developed as part of the L2B vegetation (L2BV) module. The E2E framework functions as a simulated reality that enabled to test and improve the algorithms. The models were further tuned and validated against campaign data using active learning methods. As a proof of concept, the prototype retrieval models were applied to both hyperspectral airborne (HyPlant) and spaceborne (PRISMA) imagery that were first resampled to CHIME band settings. Among the provided vegetation products, it led to a first space-based canopy nitrogen content map over a heterogeneous landscape. The obtained CHIME-like L2BV traits maps demonstrate the feasibility to routinely deliver a collection of next-generation vegetation products across the globe.

Timely and accurate information on crop growth and seasonal dynamics are increasingly needed to develop monitoring systems aimed to detect seasonal anomalies, support site specific management and estimate crop yield at the end of the season. In particular, frequent decametric information nowadays being provided exploiting the new generation of Earth Observation (EO) platforms are fundamental for farm level monitoring. This study presents an analysis aimed at fully exploiting dense time series of EO data derived from the combined use of ESA Sentinel-2A and NASA Landsat-7/8 imageries for crop phenological monitoring. Decametric Leaf Area Index (LAI) maps were generated for the year 2016 by inversion of the PROSAIL radiative transfer model with Gaussian process regression for a study area located in northern of Italy, and used to derive rice crop stages information by exploiting an adaptation of phenological algorithm. Preliminary results showed that occurrence of the main crop development stages can be estimated at parcel level with reasonable accuracy (7 to 15 days mean error depending on the phenological phase) and phenometrics analysis provide information in agreement with the different cultivated varieties and adopted agro-management practices.

The present short paper describes the functionalities of a desktop tool enabling non-expert users to develop downstream services for the periodic download and processing of data from Sentinel sources.

The agricultural sector is facing important global challenges due to the pressure of food demand, increased price-competition produced by market globalization and food price volatility (G20 Agriculture Action Plan), and the necessity of more environmentally and economically sustainable farming. Earth Observation (EO) systems can significantly contribute to these topics by providing reliable real time information on crop distribution, status and seasonal dynamics. ERMES FP7 project aims to create added-value information for the rice agro-sector by integrating EO-products in crop models. Time series of moderate resolution satellite data are analyzed exploiting the PhenoRice algorithm to retrieve seasonal occurrence of agro-practices and phenological stages. Eleven years (2003-2013) of rice seasonal metrics were derived and used in WARM crop model to set up a crop forecasting systems, with the aim to provide crop yield estimates for regional authorities. Preliminary test conducted in Italy on indica rice ecotype demonstrated that the system can provide rice yield estimates explaining up to 90% of interannual variability.

Leaf area index (LAI) is a key biophysical parameter used to determine foliage cover and crop growth in environmental studies in order to assess crop yield. LAI estimates can be classified as direct or indirect methods. Direct methods are destructive, time consuming, and difficult to apply over large fields. Indirect methods are non-destructive and cost-effective due to its portability, accuracy and repeatability. In this study, we compare indirect LAI estimates acquired from two classical instruments such as LAI-2000 and digital cameras for hemispherical photography, with LAI estimates acquired with a smart app (PocketLAI) installed on a mobile smartphone. In this work it is shown that LAI estimates obtained with the classical instruments and with a smartphone are well correlated. Consequently, results presented in this work allow considering PocketLAI as a powerful alternative to the classical instruments for LAI monitoring during field campaigns.

This paper describes a rice mapping and growth monitoring project carried out using both optical and SAR data on an agricultural land area in northern Italy. The approach implemented for mapping rice area is based on synthetic features derived from both the optical and SAR C-band multi-temporal dataset and a rule-based algorithm applied on a pixel basis. SAR X-band data were used improving winter crops recognition. Seasonal dynamics were used to identify rice growing season for regression analysis between SAR backscatter and vegetation parameters. Rice green LAI maps have been produced using the equation found during this analysis between C-band HH pol. backscatter and rice LAI.

Nowadays Spatial Data Infrastructures are the best practice to publish huge amount of spatial data on the Web in an interoperable and distributed way. Nevertheless, this operation requires a significant effort, expertise and motivation to data providers. In this paper, we propose an original approach to support geo-data providers by automating the workflows for publishing geo-data and relative web services for a given application. The prototypal solution has been tested on a real case study to support the regional or national agricultural sector in Italy.

Crop monitoring services require the capability to investigate vegetation condition, especially in regions periodically affected by low or erratic rainfalls as recently happened in northern Italy. This paper define an operational methodology based on the “triangle method” to estimate Evaporative Fraction as an indicator of surface moisture condition. The approach, that relies on low resolution MODIS products and air temperature maps, it was applied to the 2010-2014 growing seasons deriving instantaneous maps to represent daily crop conditions. The qualitative assessment of the results shows the agronomical coherence of the estimated variable, and the preliminary multi-annual analyses indicates the approach as a promising tool to support near real time crop monitoring at regional scale.

Hyperspectral remote sensing can support exploration activities helping in the recognition of lithologycal outcrops over extended areas. This is mainly due to the nature of the mineral reflectance spectra which feature very narrow absorption peaks, therefore requiring hyperspectral sensors to be assessed. In this work, the capabilities for geolithology recognition of aerial hyperspectral data have been tested to map outcrops in an Alpine carbonates test basin, located in the North-Eastern part of Italy. Two hyperspectral datasets, from MIVIS and SASI sensors, have been classified to retrieve the geolithological map of the test basin. The classification procedure was based on a decision tree approach, in order to combine different spectral and morphological features, in a flexible, robust and scalable solution. Despite the geological complexity of the study area (Alpine environment, spread vegetation cover, strong presence of shadow areas and presence of transitional geologic formations) the results, as compared to reference geological maps, prove satisfactory mapping capabilities, particularly for carbonatic rocks. In fact, the use of aerial hyperspectral data not only allows their accurate detection, but also the distinction between dolostones and limestones. The main classification issues are related to the outcrops' extent which, in a such lithology-fragmented area, avoids the detection of many small outcrops.

The application of an integrated monitoring tool to assess and understand the effects of annually occurring forest fires is presented, with special emphasis to Mediterranean and Temperate Continental zones of Europe. The distinctive features of the information conveyed by optical and microwave remote sensing data have been firstly investigated, and pertinent information have been subsequently combined to identify burned areas at the regional scale. We therefore propose a fuzzy-based multisource framework for burned area mapping, in order to overcome the limitations inherent to the use of only optical data (which can be severely affected by cloud cover or include low albedo surface targets). The relevant experimental validation has been carried out on an extensive area, thus quantitatively demonstrating how our approach successes in identifying areas affected by fires. Furthermore, the proposed methodological framework can also be profitably applied to Sentinel (optical and SAR) data.

This paper describes a mapping project carried out using both optical and SAR data on an agricultural area in northern Italy where the main crops are corn, rice and wheat. Temporal trends of backscatter and reflectance, given by the variations in vegetation growth, soil conditions and agricultural practices were analyzed and interpreted thanks to the ground-measured data. Information extracted from both optical and SAR data (vegetation indices, backscatter and texture features) were used to create training sets for implementing three different classification approaches. The work aimed at comparing early crop maps with maps derived at the end of the season. Results show that the classification accuracy obtained using only multispectral optical data is higher than the one reached using only SAR as input. Integrating both optical and SAR multitemporal features provides some advantages in terms of a more reliable crop map, especially during an early temporal stage scenario. Among the supervised algorithms tested, Maximum Likelihood shows the best overall accuracy performances at each thematic level, time step and using both optical and SAR input data.

The drought episode that struck Middle East countries in 2007-2009 was the worst one hitting the region in more than 60 years. An analysis of rainfall and vegetation dynamics over the temporal range covering 11 hydrological seasons from September 2002 to August 2013 has been carried out by using time series of satellite data from TRMM and MODIS platforms, over the study area covering the so called Fertile Crescent: spanning from southern Turkey, Syria, and Jordan, to Iraq and western Iran. The results of this analysis, along with supportive information about surface water reserves drawn from satellite radar altimetry suggest that the intensification of activities in new agricultural areas in southern Turkey, steadily developed during the last decade, contributed to the sensible deplenishment of water resources available over the basin.

A straightforward way to map burned areas from remotely sensed imagery is to integrate partial evidence of burn provided by multiple spectral indices (SIs). Our approach relies on fuzzy set theory to generate integrated layers of overall positive evidence (PE) and negative evidence (NE) scores. In order to reduce commission errors, we propose the use of NE for revising the overall PE. Revised layers are input for a region growing algorithm to produce a map of burned areas. Thematic Mapper (TM) images, acquired over the Mediterranean area, were used to derive the SIs and to define the soft constraints (membership functions). The performance of the revision process is tested for a TM image acquired over Portugal: The revision decreases the commission error from 59.5% to 1.3% and increases the overall accuracy from 42.6% up to 91.3%.

The methodology proposed in this study allows comparing different regions and analysing a national, sub-national and regional synthesis of land conditions. The synthetic information on the rainfall and vegetation trends and current states aggregated at administrative scale are crucial for governmental institutions and decision makers at various levels to support the processes of the implementation of their environmental policies, as decided in the framework of NARMA/Geoland-2 project. The results of this study can be easily incorporated in Country Environmental Profile documents, based on which the European Commission services analyse the overall situation of each country of sub-Saharan Arica.

Geologic applications of remote sensing data often rely on ancillary and support information for effectively mapping geolithologies. This work aims to investigate the capabilities of mapping geologic outcrops of sedimentary rocks using only spectral information coming from aerial hyperspectral data. MI VIS hyperspectral data obtained in the area of Serra di Falco, in southern Italy, have been exploited for testing and comparing geologic maps resulting from 4 different spectral supervised algorithms (SVM, SAM, SID, MAXLIKE) in combination with 4 different methods of selecting the training sample for feeding the classifiers, and making use of various ancillary data (ground surveys, geological map) only as reference information for evaluation the results. Geologic mapping results comparison shows the pros and cons of spectral classification in a complex sedimentary geology context.

Burned-area mapping algorithms developed for satellite images often rely on the use of spectral indices for discriminating between burns and other surfaces. The choice of the most suitable index is often a difficult task because each index brings a different type of information, as well as a rate of misclassification error. Moreover, the choice may be a function of the geographical area, spectral and geometrical characteristics of satellite data, and objectives of the study. In this letter, we compare the performance of different indices computed for Advanced Spaceborne Thermal Emission and Reflection Radiometer imagery and propose a methodology for integrating them into a synthetic indicator of likelihood of burn. The methodology is based on fuzzy set theory and aims to lay the foundation for the development of a burned-area mapping algorithm in the Mediterranean environment of southern Italy.

The monitoring of the environment's status at continental scale involves the integration of information derived by the analysis of multiple, complex, multidisciplinary, and large‐scale phenomena. Thus, there is a need to define synthetic Environmental Indicators (EIs) that concisely represent these phenomena in a manner suitable for decision‐making. This research proposes a flexible system to define EIs based on a soft fusion of contributing environmental factors derived from multi‐source spatial data (mainly Earth Observation data). The flexibility is twofold: the EI can be customized based on the available data, and the system is able to cope with a lack of expert knowledge. The proposal allows a soft quantifier‐guided fusion strategy to be defined, as specified by the user through a linguistic quantifier such as ‘most of’. The linguistic quantifiers are implemented as Ordered Weighted Averaging operators. The proposed approach is applied in a case study to demonstrate the periodical computation of anomaly indicators of the environmental status of Africa, based on a 7‐year time series of dekadal Earth Observation datasets. Different experiments have been carried out on the same data to demonstrate the flexibility and robustness of the proposed method.