#	Pagina
attuale pagina	/open-h2020/projects/206330/results.html
-1	/open-h2020/per-topic/trusting/list/index.html
-2	/open-dormire/piemonte/TORINO/id-102028/airport.html
-3	/open-h2020/per-topic/kinglist/list/index.html
-4	/open-h2020/per-country/nl/audion+therapeutics+bv/index.html
-5	/open-h2020/per-topic/neurocomputational/list/index.html
-6	/open-h2020/per-topic/prolongations/list/index.html
-7	/open-h2020/per-topic/telecollaborative/list/index.html
-8	/open-h2020/per-topic/poxvirus/list/index.html
-9	/open-h2020/per-topic/copter/list/index.html
-10	/open-dormire/piemonte/ALESSANDRIA/id-600689/la-locanda-degli-asinelli.html

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - SENSAGRI (Sentinels Synergy for Agriculture)

Teaser

Summary

Sentinels Synergy for Agriculture (SENSAGRI) aims at providing Copernicus, the European Union Earth Observation programme, with prototypes for new core services targeted to the agricultural sector.
Copernicus is an ambitious programme that is producing a wealth of information from a constellation of different satellites, the Sentinels, on a free and open data policy basis. While some useful services for the agricultural sector are already implemented, there is an enormous potential for new applications and services. Two of the Copernicus satellite missions, namely Sentinel-1 (S1) and Sentinel-2 (S2), are especially suited for agriculture applications. Both satellites obtain images at high spatial resolution (up to 10 m) with high frequency (each five days or less) and are therefore capable of providing highly dynamic field-level information at a pan-European scale.
In response of the Horizon 2020 work programme topic EO-3-2016 (Evolution of Copernicus Services), SENSAGRI aims to exploit the unprecedented capacity of S1 and S2 to develop an innovative portfolio of prototypes for agricultural monitoring services. SENSAGRI will exploit the synergy of optical (S1) and radar (S2) measurements to develop three prototype services capable of near real time operations: (1) Surface Soil Moisture (SSM), (2) green and brown leaf area index (LAI) and (3) crop type mapping. In addition, SENSAGRI proposes four advanced proof-of-concept services: (i) yield/biomass, (ii) tillage change, (iii) irrigation and (iv) advanced crop maps.
The developed algorithms will be validated in four European agricultural test areas in Spain, France, Italy and Poland, which are representative of the European crop diversity, and their usefulness demonstrated in at least three other sites in Ukraine, South Africa and Argentina. To ensure that the prototypes are developed fulfilling the Copernicus Land strategic goals, a fluent communication with the Copernicus Entrusted Entities will be kept throughout the project life. A continuous relationship with other similar projects will be also maintained, to maximize synergies and complementarities.

The main SENSAGRI project objectives are to:
1. Combine the Copernicus S1 radar with S2 optical and in-situ data in order to develop new applications and market opportunities for the European agricultural sector
2. Develop prototype services of SSM, green and brown LAI and seasonal crop type mapping and use those for proof-of-concept services of advanced agricultural monitoring products
3. Validate delivered services and establish service demonstration cases to show the large application potential of the new upstream data products
4. Disseminate prototype and proof-of-concept services and interact services with the agricultural sector and the Copernicus Entrusted Entities.

Work performed

During the first year of the Project, the work was centered in preparing a first version of the prototypes and proof-of-concept services, which required the definition and calibration of the underlying algorithms, as well as the implementation of processing chains for Sentinel-1 and Sentinel-2, to obtain preliminary products over test sites.

In parallel, the project partners agreed on the measurement procedures for all the foreseen products, on the sampling strategy for validation and on the error metrics. All the required instrumentation and methods were set up in the test sites and the first campaigns were carried out in the four core sites.

To ensure an efficient data sharing and storage, a data, image and service product repository was implemented. The partners uploaded the products generated in 2017 as well as information collected in previous years over the test sites.

The initial version of the Surface Soil Moisture (SSM) retrieval algorithm, exploiting S1 and S2 data, was developed from the SMOSAR (Soil MOisture retrieval from multi-temporal SAR dataâ€) processor. Its application in three test sites demonstrated the feasibility for obtaining SSM maps at an unprecedented spatial resolution of up to 100 m on a weekly basis.

Novel Green and Brown Leaf Area Index (LAI) retrieval algorithms, were trained and optimized through a machine learning Gaussian processes regression (GPR). LAI green/brown RGB composite maps were generated for the test sites, including bands for mean estimates, standard deviation and coefficient of variation. In addition, a multi-temporal analysis of LAI retrieval was performed in a test site in Spain.

The first version of the Seasonal Crop Mapping Service, for Pan-European crop mapping using very limited training data, was prepared and two seasonal products were defined: (1) a binary cropland mask and (2) a crop type map.

Methods and algorithms for providing advanced, added-values products were developed by relying on the synergetic use of S1-S2 data. Specifically, conducted activities were: (1) the development of a method for detecting irrigated agricultural areas by using time series of S1 SSM maps, based on the SSM contrast existing between irrigated fields and the surrounding area; (2) the development of a method to obtain binary masks of tillage changes derived from S1 and S2 temporal data; (3) the description and recoding in Python of the SAFY-CO2 model for yield forecasting and (4) the definition of the initial methodology for the added-value crop map from S1 and S2 images (ITACyL), based on machine learning algorithms and external databases; as well as the release of the 2016 added-value crop map for Castile and Leon (Spain) and a provisional version for 2017.

Final results

SENSAGRI aims at overcoming the limitations of current Copernicus products for agricultural applications, which are often unsuitable for most European agricultural areas, due to its coarse spatial resolution, its low temporal frequency or the mismatch between its delivery time and the moment when the products are more useful for stakeholders.
Figure 2 shows the expected resolutions of SENSAGRI services, compared to what is now available in Copernicus.
The services proposed in SENSAGRI could provide high-spatial resolution agricultural information suitable for field-scale monitoring applications at a Pan-European scale. They could feed a variety of information and Decision Support Systems, operated either by public or private organizations, serving as a basis for new downstream Copernicus services in the agribusiness.