Volume 73, Issue 6 e13315
DATA ARTICLE
Open Access

European Soil Data Centre 2.0: Soil data and knowledge in support of the EU policies

Panos Panagos

Corresponding Author

Panos Panagos

European Commission, Joint Research Centre (JRC), Ispra, Italy

Correspondence

Panos Panagos, European Commission, Joint Research Centre (JRC), Ispra, Italy.

Email: [email protected]

Contribution: Conceptualization, ​Investigation, Writing - original draft, Methodology, Validation, Visualization, Writing - review & editing, Formal analysis, Project administration, Data curation, Supervision, Resources

Search for more papers by this author
Marc Van Liedekerke

Marc Van Liedekerke

European Commission, Joint Research Centre (JRC), Ispra, Italy

Contribution: Conceptualization, Supervision, Writing - review & editing

Search for more papers by this author
Pasquale Borrelli

Pasquale Borrelli

Department of Science, Roma Tre University, Rome, Italy

Contribution: Writing - original draft, Methodology, Visualization, Writing - review & editing, Data curation

Search for more papers by this author
Julia Köninger

Julia Köninger

European Commission, Joint Research Centre (JRC), Ispra, Italy

Contribution: Methodology, Formal analysis, Data curation, Visualization

Search for more papers by this author
Cristiano Ballabio

Cristiano Ballabio

European Commission, Joint Research Centre (JRC), Ispra, Italy

Contribution: Conceptualization, ​Investigation, Visualization, Validation, Formal analysis, Software, Data curation

Search for more papers by this author
Alberto Orgiazzi

Alberto Orgiazzi

European Commission, Joint Research Centre (JRC), Ispra, Italy

Contribution: Conceptualization, Formal analysis, Supervision, Data curation, Writing - review & editing, Writing - original draft

Search for more papers by this author
Emanuele Lugato

Emanuele Lugato

European Commission, Joint Research Centre (JRC), Ispra, Italy

Contribution: Writing - original draft, ​Investigation, Writing - review & editing, Formal analysis

Search for more papers by this author
Leonidas Liakos

Leonidas Liakos

European Commission, Joint Research Centre (JRC), Ispra, Italy

Contribution: Software, Formal analysis, Data curation, Validation, Methodology, Visualization

Search for more papers by this author
Javier Hervas

Javier Hervas

European Commission, Joint Research Centre (JRC), Ispra, Italy

Contribution: Writing - review & editing, Writing - original draft

Search for more papers by this author
Arwyn Jones

Arwyn Jones

European Commission, Joint Research Centre (JRC), Ispra, Italy

Contribution: Project administration, Supervision, Funding acquisition

Search for more papers by this author
Luca Montanarella

Luca Montanarella

European Commission, Joint Research Centre (JRC), Ispra, Italy

Contribution: Project administration, Supervision, Funding acquisition, Conceptualization

Search for more papers by this author
First published: 13 October 2022
Citations: 12

Abstract

The European Soil Data Centre (ESDAC), hosted by the European Commission's Joint Research Centre (JRC), is the focal point for soil data, support to policy making and awareness raising for the European Union (EU). Established in 2006 to provide harmonised soil-related information for the EU Member States, and ESDAC currently hosts 88 datasets, 6000 maps, six atlases, 500 scientific publications, and a copious amount of soil-related material. Through its data repository publishing activity, ESDAC has licensed over 50,000 datasets during the past 15 years; 8500 of them in 2021 alone. It has published 140 monthly newsletters and is followed by more than 12,000 subscribed users, which receive regular updates. This article addresses the use, usability, and usefulness of ESDAC. About 75% of the ESDAC users come from academia and the research community while the remaining 25% includes public administration (at EU, national, regional, and local level) and the private sector. In addition, we provide some insights of the datasets evaluation and how they have been developed. The general ESDAC vision is to provide evidence underlying EU soil-relevant policies and to facilitate the access to relevant data for research. ESDAC is an integral part of the recently established European Union Soil Observatory (EUSO), with a target to have an even stronger role in supporting EU and regional policies.

Highlights

  • ESDAC is a central place from where to find European wide relevant soil data.
  • ESDAC is an integral part of the EU Soil Observatory in creating data and knowledge for policy support.
  • The ESDAC website shows a high volume of traffic; 10,000 of user licenses are granted per year.
  • ESDAC key success: open access data policy, documentation, operational helpdesk and regular updates.

1 INTRODUCTION

The European Commission's (EC) vision is to be the first climate-neutral continent by 2050 (Čavoški, 2020). To achieve this major challenge, the EC presented the European Green Deal, the most ambitious package of measures that should enable European citizens and businesses to benefit from sustainable green transition. In the past two years, the EC has proposed ambitious objectives through the Soil Strategy 2030, the Biodiversity Strategy 2030, the Farm to Fork, the Zero Pollution Action Plan, and the European Climate Law, which include actions for sustainable soil management.

In December 2020, the EC launched the European Soil Observatory (EUSO) to better support EU environmental, agricultural, and climate policies, and to be the long-term central scientific and technical point of reference in the EC concerning all soil matters. One of the five objectives of the EUSO is to expand the European Soil Data Centre (ESDAC) and reinforce its data sharing capacity (Figure 1), by contributing to EUSO's four other objectives: (i) establishment, operation and optimisation of an EU wide soil monitoring system, (ii) development of a dashboard of soil indicators to support EU policies, (iii) development of an open and inclusive discussion forum of stakeholders (decision makers, public administration, private sector, etc.), and (iv) support to regional activities on soil research and innovation (Panagos et al., 2022).

Details are in the caption following the image
European Soil Data Centre (ESDAC) 2.0 and EU soil observatory

ESDAC (https://esdac.jrc.ec.europa.eu/) is an online web platform hosting a series of pan-European and global datasets, maps and soil-related documents. Originally, ESDAC was an initiative of the EC with the purpose of being the reference point for soil data and associated knowledge at European level, not only for internal use by the EC and associated European institutions, but also as a service to European citizens. Since its launch in 2006, ESDAC has grown in an ad-hoc fashion, and has tried to include the latest state-of-the-art know how in soils at pan-European scale.

The core business of ESDAC is to make the data from its soil data repository accessible, augmented by providing access to a large number of maps, atlases, publications, and technical reports. In addition, ESDAC has a knowledge base section with detailed information on soil threats and other themes, networks, and a selection of EU-funded soil projects.

ESDAC has its place among soil data centres around the world. Examples are: ISRIC (https://www.isric.org/)-World Soil Information Service (WoSIS) has as mission to serve the international community as custodian of global soil information (Ribeiro et al., 2018). In the United States, the US Department of Agriculture (USDA) and the National Resources Conservation Service (NRCS) (https://www.nrcs.usda.gov/wps/portal/nrcs/site/soils/home/) are the main focal points for US soil surveys (Schaefer et al., 2007). In Australia, CSIRO (https://www.csiro.au/en/research/natural-environment/land/soil-and-landscape-grid-of-australia) has collected and organised the national soil data (Johnston et al., 2003; Minasny & McBratney, 2001). In Europe, some countries operate a national soil data service. Excellent examples are the French National Soil Data Centre (Richer-De-Forges & Arrouays, 2010) and the BonaRes data repository in Germany (Grosse et al., 2020).

The scope of the ESDAC data is its use for policy development, implementation, and assessment in a number of policy areas both at EU and national scale. The data cover soil properties, soil functions, and soil threats. The policy areas include agriculture, soil protection, sustainable development, bio-energy, water protection, and nature protection. Within the new EU Green Deal, soil becomes a crucial element, contributing to EU climate neutrality ambition in 2050 (soils as carbon pool), a zero pollution ambition for toxic-free environment (soil contamination), preserve and restore ecosystems and biodiversity (soil biodiversity), environmental friendly food system and Farm to Fork (nutrient management, pesticides). Examples on how ESDAC data support the EU policies can be found in the Supplementary material.

The objective of this article is to (a) give an overview of the datasets in ESDAC; (b) demonstrate the impact of ESDAC in research, policy, private sector and awareness raising; (c) evaluate the type of users and the main usages of ESDAC data, and (d) propose new developments in line with the implementation of the new EU soil-relevant policies and (e) the establishment of the EU Soil Observatory.

2 ESDAC TECHNICAL SET UP

Being an integral part of the EUSO, it is required that ESDAC will have a long-term operability. Therefore, the maintenance and further software development is ensured in the long term. Technically, ESDAC has been developed in the Drupal content management system environment (Patel et al., 2011). The back-end of ESDAC uses the internal database of Drupal content management system to store the website interface. The data requests are stored in an external database in PostgreSQL database that is an open-source Relational Database Management System (RDBMS).

ESDAC operates a help desk that cares for the timely assistance for any matter related to ESDAC use and its contents, the downloading of data, the notification of soil-data-related news as well as the scientific and technical support for the use of the various products. Users can ask questions and raise issues through the use of the ESDAC functional mailbox: [email protected]. In addition, a powerful search engine may help the user to find the requested information (e.g., searching with keyword “Ukraine” returns 13 records).

The ESDAC data are distributed to users with the Hypertext Transfer Protocol Secure (HTTPs) which ensures that the data transfer is safe and is only accessible to the person who is authenticated to view that data (Shah & Correia, 2021). Users can request data through a simple web interface after registration with their details (name, e-mail, country, and affiliation) and the reason for requesting such data, after which they receive an e-mail with details on how to access the data (valid for 15 days). Most ESDAC data can be used for any purpose with the only limitation not to distribute the data to third parties and with the requirement to cite the data properly when used in publications or data-products. The registration serves only for statistical purposes (see next chapters). All datasets are free to download and they are listed in the web address: https://esdac.jrc.ec.europa.eu/resource-type/datasets.

3 ESDAC DATA SUMMARY

In the supplementary material (Table S1), we present an overview of the 73 available datasets grouped in five categories that address (a) European soil database (ESDB) and World Reference Base (WRB), (b) point data and soil properties, (c) soil threats, (d) soil functions, and (e) global data. An additional group of 15 secondary datasets (Table S2 in Supplementary material) is also available in ESDAC. The secondary datasets are either limited to specific regions or countries, published in other repositories or are results of meta-analysis; therefore, the registration process is not requested.

In the following sub-sections, we briefly describe the most user-requested datasets. Statistics refer to the 8,428 data requests of the year 2021 (Table S3 in Supplementary material) in order to allow a fair comparison between old and new datasets. The figures in this section (Figures 2-4) are a graphic representation of selected datasets for illustration purposes. The high resolution maps and data are available from the corresponding ESDAC pages.

Details are in the caption following the image
European soil database (ESDB) (a); soil physical property maps (b–e); chemical property maps (f–k)
Details are in the caption following the image
Examples of soil threats datasets: Erosion processes (a–e), soil organic carbon (f–i), landslides (j), biodiversity (k–m), and diffuse soil pollution (n–o)
Details are in the caption following the image
Global datasets available in ESDAC

3.1 European soil database (ESDB) and derived products

The ESDB (Figure 2a) is ESDAC's flagship product and provides EU-wide data for 73 soil attributes; it is available since 2006 and it is the most requested dataset (15% of the total data requests in 2021). The ESDB is the result of a collaborative project involving all authoritative soil organisations in the European Union and neighbouring countries in the late 1990's. At a scale of 1:1 million, the ESDB is a simplified representation of the soil coverage diversity and spatial variability in Eurasia. Up to the present day, it is the only harmonised soil database at European scale using a participatory approach with experts from all countries. As ESDB is available in two formats, 2/3 of the users request the vector version (polygon based) of the database while 1/3 prefer the raster version.

The ESDB includes 73 attributes addressing soil classification (WRB, FAO), texture, parent material, land use, impermeable layer, limitation to agricultural use, obstacle to roots, soil water regime, altitude, slope and physical, chemical, mechanical, and hydrological properties (King et al., 1994). It should be noted that the properties are not expressed as numerical values, but as categorical values.

3.2 Land Use and Cover Survey (LUCAS) point data and soil physical/chemical properties

The LUCAS soil module is part of the LUCAS that aims to improve the pan-European land cover mapping (Pflugmacher et al., 2019) and better describe land cover and use diversity at regional level (d'Andrimont et al., 2020). The LUCAS soil module entails soil sampling at a large number of points across EU (more then 20,000), consecutive soil analysis, and sample storage. Such campaigns took place in 2009, 2015, and 2018, and are repeated in 2022. The LUCAS soil datasets include data for particle size distribution (silt, clay, and sand), coarse fragments, soil pH, organic carbon content, carbonate content, total nitrogen, phosphorus, potassium, and cation exchange capacity (Orgiazzi et al., 2018) (Figure 2 b-k). LUCAS data are made available for 2009 (Tóth, Jones, & Montanarella, 2013) and 2015 (Jones et al., 2020; Orgiazzi et al., 2018), and 2018. The LUCAS topsoil datasets are among the most requested, recording 11.6% of the total requests in 2021. In addition to the 13 measured physical and chemical properties, LUCAS 2009 and 2015 include a corresponding visible and near infrared spectral library (Nocita et al., 2014).

The LUCAS topsoil database were used to map physical (Ballabio et al., 2016) and chemical soil properties (Ballabio et al., 2019) for the geographical extent of EU (Figure 2b–k) using advanced modelling techniques such as Gaussian Process Regression (GPR) or Multivariate Adaptive Regression Splines (MARS). LUCAS topsoil database have also contributed to the development of other global maps of soil properties such as SoilGrids (Hengl et al., 2017; Poggio et al., 2021).

In 2021, the data requests for LUCAS point data, derived physical and chemical attributes, and the soil profile data summed up to 20% of the total ESDAC data requests (Figure S1).

3.3 Soil threats

With 30 datasets addressing soil threats, this is the major category in ESDAC data repository. It reflects 25.9% of total data requests in 2021, which is the highest among all categories (Table S3). Soil compaction and soil salinisation are among the less-addressed soil threats where updated pan-European assessments are missing.

3.3.1 Soil erosion

ESDAC includes quantitative and qualitative pan-European assessments of soil loss by water, wind, gully, tillage, and (root) crop harvest erosion (Figure 3a–d). Most of the available erosion datasets deal with water erosion, for which different modelling approaches have been used. These include WaTEM/SEDEM (Borrelli et al., 2018), RUSLE (for the years 2010, 2016, 2050) (Panagos et al., 2021), PESERA (Kirkby et al., 2008), and MESALES (Le Bissonnais et al., 2002). The RUSLE-based geo-statistical approach allowed for the first time to account the effects of the different regional cropping systems and their management practices into a pan-European soil erosion model (Panagos et al., 2015). This provided unique insights into the potential mitigating effects attributable to conservation agriculture across the EU. In a following step, these modelling improvements were integrated in the WaTEM/SEDEM model (Borrelli et al., 2018), which provided for the first time a pan-European estimate of sediment yields and sediment supplies to the riverine system.

Concerning wind erosion, three pan-European datasets are available in ESDAC, (i) a quantitative estimate of soil loss by wind erosion for arable lands (Borrelli et al., 2017a); (ii) a fuzzy-logic-based qualitative estimate of land susceptibility to wind erosion (Borrelli et al., 2016); and (iii) wind erosion erodibility.

ESDAC is also providing datasets on (i) the first pan-European assessments of other geomorphic processes such as soil loss due to (root) crop harvesting (e.g., sugar beets and potatoes) (Panagos et al., 2019); and (ii) recent developments in gully erosion research, such as the national scale gully erosion inventory (Vanmaercke et al., 2021).

Among the soil threats, the group of 13 soil erosion datasets are the most requested, covering 15.5% of total data requests; this proportion rises to 28.5% if we add the global erosion datasets.

3.3.2 Soil organic carbon

ESDAC hosts three types of datasets on soil organic carbon (SOC) (Figure 3f–i): (a) use of pedotransfer rules such as OCTOP (Jones, Hiederer, et al., 2005) or aggregated data, (b) geo-statistical methods applied to LUCAS 2009 datasets to produce SOC concentration maps (de Brogniez et al., 2015; Yigini & Panagos, 2016), and (c) advanced biogeochemical models such as CENTURY to estimate carbon stocks and fluxes (Lugato, Panagos, et al., 2014). The latter detailed modelling framework, driven by state-of-art soil datasets from ESDAC and statistical data (EUROSTAT, FAO), was used to make scenarios analysis on best practices to sequester soil carbon (Lugato, Bampa, et al., 2014). The modelling framework was then extended to include lateral C fluxes by coupling CENTURY with the RUSLE-based approach described above. By some parsimonious assumptions on sediment displacement, the integrated modelling was the first attempt to include feedbacks on C cycle related to erosion and deposition processes at the continental scale, in combination also with future climate change scenarios (Lugato et al., 2018).

The model framework was successively improved adopting the daily time-step model DayCent, first run in LUCAS points (Lugato et al., 2017) and then at 1 km grid. This allowed to make more inclusive scenarios on potential mitigation strategies, taking into account the interactions between C and N cycles and the associated GHG fluxes.

Data-driven approaches, based on LUCAS soil organic matter fractionations, allowed the investigation of soil organic carbon formation and stabilisation mechanisms (Cotrufo et al., 2019) and the first continental mapping of Particulate Organic Matter (POM) and Mineral-Associated Organic Matter (MAOM) available in ESDAC (Lugato et al., 2021). The group of seven datasets addressing SOC has a 4.5% proportion of total user requests.

3.3.3 Landslides

ESDAC is also the dissemination platform of the collaborative data produced by the JRC-hosted European Landslide Expert Group, which was created in 2007 to identify areas at risk of landslides in Europe (Hervás et al., 2007).

The ESDAC repository currently hosts version 2 of the European Landslide Susceptibility Map (ELSUS v2). The map (Figure 3j) shows five levels of spatial probability of generic landslide occurrence at a resolution of 200 m covering all EU Member States and several neighbouring countries (Günther et al., 2014; Wilde et al., 2018). ELSUS v2 (Figure 3j) has been produced by regionalising the study area based on elevation and climatic conditions, followed by spatial multi-criteria evaluation modelling using slope angle, shallow sub-surface lithology, and land cover spatial datasets as the main landslide conditioning factors. In addition, the location of over 149,000 landslides across Europe, provided by various national organisations or collected by the authors, has been used for model calibration and map validation. The database also include ancillary maps used for landslide susceptibility modelling and a confidence level map of ELSUS v2 on Eurostat NUTS 3 regions level.

3.3.4 Soil biodiversity

ESDAC includes information and datasets related to the biological soil component (Figure 3k–m). Due to its increasing relevance, efforts have been taken over the years to expand data collection on soil life at an EU level. Both indirect assessments (e.g., potential risk) and direct measurements (e.g., microbial biomass) are available. A first attempt to assess threats to soil biodiversity took into account 13 potential stressors to soil organisms. The level of risk and its spatial distribution for the EU was determined for three major components of soil life: microorganisms (Figure 3k), fauna, and biological functions (Orgiazzi, Panagos, et al., 2016). A similar approach was used to develop a preliminary distribution of risk to soil organisms at global scale (dataset available on ESDAC as part of Global Soil Biodiversity Atlas Maps).

As of 2018, the LUCAS soil module included a soil biodiversity component to investigate soil taxonomical and functional diversity by means of metabarcoding and metagenomics in 1000 locations across the EU (Orgiazzi et al., 2022). Four targets were chosen for taxonomical analyses: archaea (16 S), bacteria (16 S), fungi (ITS), and other eukaryotes (18 S). In 2023, ESDAC plans to release a large set of soil biological datasets and derived maps based that will underpin the first assessment of soil biodiversity for the EU. Collaborations between JRC and European research institutions have further expanded the biological parameters that were measured from the LUCAS 2018 soil biodiversity samples. For instance, maps of soil microbial biomass and respiration for the EU were created and made available through ESDAC (Smith et al., 2021).

3.3.5 Soil diffuse pollution

The available data on diffuse soil contamination include maps of metal concentrations using the FOREGS Geochemical database (Lado et al., 2008) and LUCAS samples (Tóth et al., 2016). In recent years, high resolution (500 m, 1 km) datasets have been made available for copper concentration (Ballabio et al., 2018), mercury stocks in topsoil (Ballabio et al., 2021), and radionuclides (Caesium, Plutonium) (Meusburger et al., 2020). From the 21,682 analysed samples of LUCAS, we developed the copper map (Figure 3o) and also investigated the main anthropogenic sources of high copper concentrations in vineyards and orchards. The assessment on mercury concentrations (Figure 3n) addressed both natural hotspots and the anthropogenic sources but also the fluxes within major river basins. The forthcoming cadmium, zinc and arsenic assessment will further give better insights into diffuse soil pollution status in EU.

3.4 Global datasets

While ESDAC's main focus is on EU policy support, during the past few years, collaboration with external partners has led to the development of several global datasets (Figure 4). These include the Global Soil Biodiversity map and Soil Biodiversity threats (Figure 4a) (Orgiazzi, Bardgett, et al., 2016), as well as the Global Soil Erosion (Figure 4e) dataset, which provide data for 2001 and 2012 (Borrelli et al., 2017b) together with the input factors data on soil erodibility, cover management, slope length, and steepness, at a resolution of 25 km. The Global phosphorus losses dataset focus on the phosphorus content and losses due to erosion processes on arable land (Figure 4i) (Alewell et al., 2020). The soil organic carbon estimates (Figure 4d) include the global organic carbon density (t ha−1) for the topsoil (0–30 cm) and the subsoil layer (30–100 cm) (Hiederer & Köchy, 2011). The global land degradation in arable lands (Figure 4b) dataset includes five land degradation processes: aridity, soil erosion by water, vegetation decline, soil salinisation, and soil organic carbon decline (Prăvălie et al., 2021). The global erosivity dataset (Figure 4c) include the first assessment of the R-factor at global scale using measured rainfall erosivity data from 3625 precipitation stations in 63 countries (Panagos et al., 2017). The land degradation debt (Figure 4g) estimates the “environmental debts” (difference between natural potential and actual) for (a) tree cover, (b) soil erosion, (c) above ground carbon, and (d) below ground carbon (Wuepper et al., 2021). The global spatial layers for estimating soil GHG emissions from Indirect Land Use Changes (ILUC) due to the Production of Biofuels (Figure 4f) include a group of datasets such as GHG emissions from changes in soil C-stocks, global land cover, ecological zones, crop surfaces, and carbon stocks (Hiederer et al., 2010). The global erosion by 2070 (Figure 4h) includes future projections (2070) for soil erosion based on land use changes and climate change effects taking into account 16 models for three Representative Concentration Pathways (RCP2.6, RCP4.5 and RCP8.5) (Borrelli et al., 2020).

The 10 global datasets in ESDAC (Table S1; 9 in Figure 4 and landform classification) represent about 21.8% of the total requests in 2021, showing the importance of such data for many users.

3.5 Soil functions

Datasets associated with soil functions became recently available; most of them have been developed during the last seven years. Ten datasets (Table S1) in ESDAC contribute to 5.1% of the user data requests (Figure S1). Soil functions data include water retention, saturated water content, saturated hydraulic conductivity, and Mualem-van Genuchten parameters (Tóth et al., 2017). Also, there is an equal request for other soil functions data such as soil biomass productivity (Tóth, Gardi, et al., 2013), preservation of cultural artefacts (Kibblewhite et al., 2015), and raw materials availability (Table S2).

3.6 Projects datasets

The soil projects datasets are a collection of datasets that stem from various collaborations in the context of EU-funded research projects. The 7th Framework Programme (FP7) and the subsequent HORIZON2020 financed many research projects on soils. JRC participated in G2 (Karydas & Panagos, 2018), iSOIL (Werban et al., 2010), and SoilTrec (Banwart, 2011) and data outputs for these FP7 projects are available to public users. However, this collection is less developed (with only 4.9% of 2021 data requests; Figure S1) due to Intellectual property right issues. However, more than ¾ of requests on projects data refer to the collaborative activities with European Food & Safety Agency (EFSA) on crop data and the PERSAM tool for predicting environmental concentrations in soils of plant protection products (Hiederer, 2012).

4 ESDAC DATA EVALUATION

The data in ESDAC are the results of monitoring campaigns (e.g., LUCAS), in-house modelling, EU funded projects, and collaborative research with other research institutions, academia, and EU national experts. The data are supported by peer review publications and reports, while the ESDAC data download pages provide the metadata that describe spatial and temporal coverage, measurement units, and references.

The general LUCAS sampling campaign is based on a regular grid defined by the intersection points of 2 km × 2 km grid covering the EU and resulting into 1000,000 sampling points (d'Andrimont et al., 2020). From this pool of points, approximately 270,000 locations are visited by surveyors every three years to observe the land cover and other landscape features (Eurostat, 2020; Martino et al., 2009). In around 10% of these visited points, a topsoil sample is taken. The selection of the soil sampling points is based on a stratified sampling scheme using land use and other terrain information (Tóth, Jones, & Montanarella, 2013). The selection of LUCAS soil surveyed points has a significant bias towards agricultural land which implies that the LUCAS datasets may over represent the heavily sampled areas. As LUCAS soil surveyed points are allocated based on the extent of the country and not the heterogeneity, this can be an additional source of uncertainty (Carre et al., 2013).

For some soil threats assessments, we have used well-known models such as RUSLE or PESERA for soil erosion by water, RWEQ for wind erosion, WaTEM/SEDEM for sediments distribution, CENTURY for soil organic carbon, and DayCent for nitrogen emissions. For some cases, the modelling outputs have been validated using the LUCAS topsoil data (in case of carbon) and verified with national and regional assessments (in the case of soil erosion). In addition, we have used machine learning techniques for developing spatial maps for physical and chemical properties. Among others, we used deep neural network (DNN) learning models (Adnan & Akbar, 2019), generalised linear model (GLM) (Quinonero-Candela et al., 2007), GPR (Rasmussen & Williams, 2005), Cubist regression model (Quinlan, 1992), and MARS (Friedman, 1991). The machine learning models were trained using remote sensing data, climatic data and topography, at different scales according to data availability and the target scale of the end product (set as 250 m for the earlier products and 100 m for the later ones). Multi-temporal spectral data (MODIS, Sentinel 2, Landsat) were used to derive covariates addressing land cover type and its seasonal change, surface temperature, and humidity. Elevation models derived from SRTM and Aster DEM were used to derive topographic indices of wetness, slope, solar radiation, topographic position, among others (see Ballabio et al., 2019 for a description). Climatic data derived from the WorldClim database were also added to the models. The produced maps of physical (Ballabio et al., 2016) and chemical properties (Ballabio et al., 2019) have associated uncertainties and other metrics such as the root mean square error (RMSE) and mean absolute error (MAE) estimated for model fitting and by cross-validation. Uncertainties were derived directly from model estimation for GPR and GLM, and by Monte-Carlo simulations for other models.

In contrast to ESDAC's openness and data sharing practices, the majority of researchers or organisations do not share their scientific data (Tenopir et al., 2011). The release of the ESDAC scientific data allows users to better compare their local, regional or national soil-related maps with the European ones. This process is beneficial in two ways: it allows users to evaluate and possibly adapt their own data holdings vis a vis similar data at European scale, and users can provide potentially feedback to ESDAC when the European data significantly deviate from local ones. In the future, the ESDAC products can be further improved with the implementation of a pan-European Soil Information system which will get on board soil sampling data from national surveys. Currently, the latter is the subject of intense debate in the context of the new EU Soil Strategy and is guided by technical work on harmonised soil monitoring at the EU Soil Observatory.

5 ESDAC PERFORMANCE

5.1 Data usage metrics

Based on the data log of users' requests, we made an elementary statistical analysis. ESDAC received around 50,000 data requests during the past 15 years (2006 – March 2022) with an increasing trend during this period. One useful metric for measuring ESDAC performance is the number of requested datasets per year (Figure 5a). In the period 2006–2010, ESDAC granted 1645 access to public users (Panagos et al., 2012). In the period 2011–2015, we notice an annual increase in the requests of ca. 30% which results in 3553 requests in 2015, three times more than in 2011 (Figure 5a). In 2021, 8431 requests were granted, which is double that of 2017 (Figure 5a); the projections for 2022 is to receive ca. 9500 data requests. The distribution per dataset can be seen in the Supplementary material (Figure S1; Table S3). Compared to the first assessment of the ESDAC in 2011, the data requests in 2021 are one order of magnitude higher. This is not only due to the increased number of datasets (10 in 2011 vs. 88 in 2021), but also due to the increased awareness of the public of ESDAC as a focal point for soil data in Europe. The red line (Figure 5b) is the cumulative number of datasets as each year we add 5–10 datasets in ESDAC repository.

Details are in the caption following the image
(a) The trend in the ESDAC distributed datasets; (b) the cumulative number of released datasets. For 2022, an estimation is provided

Following the current trends in ESDAC requests (+12% annual increase), we expect more than 27,000 data requests per year in 2030. In addition, the available datasets are projected to double as we expect new data flows from Earth Observation systems, innovative modules in LUCAS (soil biodiversity, pesticides, microplastics, and PFAS), and EU research framework programmes (HORIZON Europe, Soil Mission projects).

The private sector is a major user (14% of total data requests) of the ESDAC data (Figure 6). More than 2600 private companies made 6700 requests during the past 10 years. These companies used the data for crop yield prediction, chemical exposure assessments, registration of plant protection products, risk assessments, land use planning, climate change studies, etc. The major requests of the private sector are for the EFSA datasets (48%) and the landslides (20%). Those two datasets are ranked as the most requested also by public administrations. The private sector was just a marginal user of ESDAC data (<5% of total data requests) 10 years ago; therefore, this geometric increase of private sector interest in soil data is expected to continue in the next decade.

Details are in the caption following the image
Users of the ESDAC datasets

Most users come from academia (56%) and research organisations (18%) (Figure 6). Academia uses the ESDAC datasets for studies, modelling, PhDs, master theses, mapping, academic assignments, teaching, papers, etc. When visually analysing, through the use of a word cloud (Supplementary material Figure S2), the purpose for which university and research users are using the data, it is shown that most requests are related to erosion, organic carbon, agriculture, soil properties, land, species distribution, and ecosystem services including also some regional trends (global, European, France, Spain, UK, etc.). The cloud for private sector and public administration users show that analysis, mapping, evaluation, planning, risk assessment, registration, and modelling are the most dominant uses (Supplementary material Figure S2).

The data users category “other” covered 4.7% of the ESDAC data requests (Figure 6). Among these, we identify individual farmers, advisory services, charities, associations of farmers, NGOs, consultants, international organisations, media, and publishers. We noticed the particular use of some ESDAC datasets (such as soil erosion, soil erodibility, and soil organic carbon) to develop a patent for a method of agronomic experimentation and fertiliser management (European Patent Office, 2019).

ESDAC seems to be quite known internationally as datasets have been requested by users in 187 countries (Figure 7). European countries (Germany, France, Italy, United Kingdom, Spain, Netherlands, Greece, Portugal, Belgium, and Romania) together with United States and China are the most frequent users (>1000 data requests) during the past 10 years. The other EU countries, together with Switzerland, India, Turkey, Canada, Brazil, South Africa, Norway, Iran, Ethiopia, Zambia, and Australia belong to the second group of users (>200 data requests). For 85 countries, we estimated requests in the range of 10–200, while 65 countries, most of them in Africa, have only a few data requests (<10). Compared to a decade ago, we noticed a geometric increase in the data requests from China, India, Turkey, Brazil, Ethiopia, Iran, and Australia. In the coming years, we expect an increasing interest from the African countries.

Details are in the caption following the image
Distribution of ESDAC datasets per country

5.2 Visits and downloads

The visits, pages viewed, and the duration of sessions are among the most popular websites metrics (Huntington et al., 2008). We used webserver statistical software (Piwik) (Chandler & Wallace, 2016) to analyse the data log of ESDAC and make a statistical analysis. ESDAC has an average of 1000 unique visitors per day, reaching almost 300,000 unique visitors and 190,000 downloads in 2021 (Figure 8a). This makes ESDAC among the top thematic scientific portals in the Joint Research Centre (JRC). Each year, there is an increase in unique visitors of around 10% compared to the year before.

Details are in the caption following the image
ESDAC visits and downloads (a); users by search engine or other websites (b); and by device type (c)

In 2021, the entry points to ESDAC included direct entries (27.9%), entries from a link of another website (8.6%), and entries through a search engine (63.4%) (Figure 8b). Direct access (direct traffic) to a website occurs when a visitor arrives directly on a website without having clicked on a link on another site (Sakas & Giannakopoulos, 2021). The ESDAC data pages appear to be very popular in Google searches. Each direct entry visit to ESDAC has a mean action of 3.6 activities per visit in almost four minutes. The search engine entries remain less (2.5 minutes) with less activity. This is also confirmed by other Google Analytics research (Plaza, 2011). The 25,132 visitors (Figure 8b) redirected from other websites have entered through FAO (14.6%), Twitter (11.8%), Facebook (6.3%), ISRIC (4.8%), pesticidesmodels.eu (3.8%), and the EUROPA website (3.6%). The remaining 55% includes hundreds of websites that contribute with less than 3% of ESDAC entries each.

An interesting finding from the 2021 data log shows that 18.5% of ESDAC visitors access through smartphones while tablets (and phablet) are used by 2.3% of users (Figure 8c). Since the global information technology trend is to use increasingly smartphones (“Going mobile”), a future ESDAC should be adapted by having content formatted accordingly.

6 KNOWLEDGE AND AWARENESS RAISING

ESDAC includes numerous documents which contribute both to soil awareness raising and soil research, which are two important objectives of the Soil Strategy and the EUSO. Major awareness raising instruments are the series of soil atlases, which describe the differences in soil patterns in different parts in an easily understandable language, in combination with many illustrations. The large number of self-standing soil maps are also a rich source of knowledge, while reports and publications contribute to research and soil assessments. Much of the knowledge in ESDAC is organised along major soil themes where users can find information about soil threats, functions, and properties. The ESDAC projects part contains information on FP7 and Horizon projects in which JRC has participated.

Since 2010, ESDAC publishes a one-page monthly newsletter with information on updates. Newsletters are sent out to a large user base of subscribers (>12,000). In addition, parts of those newsletters are disseminated through social media and other newsletters (e.g., International Union of Soil Societies).

6.1 Atlases

ESDAC has published six atlases in the period 2005–2016 (Figure 9). Those are the results of fruitful collaborations with researchers and institutes worldwide. The Atlases contribute to raising awareness about soils. Data coming from different countries and sources were processed to produce harmonised maps; they indicate solutions for sustainable management of soils and provide the ground to build strong collaborations worldwide.

Details are in the caption following the image
Covers of the published atlases in ESDAC. On top of each picture: The year of publication and the languages in which the atlas was published (English, French, Spanish, and Portuguese)

The Soil Atlas of Europe (Jones, Montanarella, et al., 2005) is the result of a collective effort by more the 40 national soil surveys and soil science institutions cooperating across Europe under the European Soil Bureau Network (ESBN). The Soil Atlas of the Northern Circumpolar Region (Jones et al., 2010) was the result of a three-year collaborative project with partners from northern EU countries, as well as Norway, Iceland, Greenland, Canada, the USA, and Russia undertaken under the auspices of the International Polar Year 2007–2008. The European Atlas of Soil Biodiversity (Jeffery et al., 2010) positively showcases inter-service collaboration between the JRC and other EU Commission services, together with productive partnerships with internationally renowned scientists from all over the world for the of International Year of Biodiversity 2010 (also available in French). The Soil Atlas of Africa (Jones et al., 2013) raised awareness of the importance of soil in Africa to the general public, policymakers, land users and scientists of other disciplines, and it provided indirect support to EU policies and instruments for Development and Aid Assistance (also available in French). Coordinated by the European Commission's Joint Research Centre (JRC), the Soil Atlas of Latin America and Caribbean (Gardi et al., 2015), published in three languages, is the result of a fruitful collaboration with leading soil scientists in Europe, Central and South America, and the Caribbean. Finally, the JRC and the Global Soil Biodiversity Initiative (GSBI) together with 70 organisations published the first-ever Global Soil Biodiversity Atlas (Orgiazzi, Bardgett, et al., 2016) that maps the soil biodiversity of the entire planet. The Soil Atlas of Asia is currently under development as is a Chinese translation of the Global Soil Biodiversity Atlas. The data for the Global Biodiversity Atlas, Soil Atlas of Africa and Northern Circumpolar Region are made available in ESDAC (Table S1).

6.2 Maps, reports, and publications

Besides the atlases, the ESDAC hosts 470 documents (produced in-house) supporting the ESDAC datasets and its knowledge base. A major part of this repository refers to publications (307 published papers) that have been made available during the past twelve years (2010–2021). The remainders are technical reports and supporting documents for the ESDB, LUCAS, and other projects covering the past 20 years.

Publications in journals include more than 307 papers from the team dealing with soil at JRC. Most papers refer to the past eight years (2013–2021). In many cases, they document the datasets published in ESDAC. About 25–30 publications are added each year in this repository.

ESDAC hosts a collection of more than 6,000 maps, including 5400 national-scale soil maps, a collection of maps derived from the ESDB and a collection of maps associated to publications. For about 40 years, ISRIC–World Soil Information has been providing significant support to the international science community by collecting and archiving regional-, national- and global-scale maps of soils and land resources (Arrouays et al., 2017). Despite effective procedures for storage and maintenance, most organisations involved in archiving struggle to arrest the deterioration of paper maps and the quality of information they contain. Realising the need to conserve the information on existing maps, which underpin the fast-developing thematic mapping strategies to support soil protection, in 2006, the JRC and ISRIC initiated the European Digital Archive of Soil Maps (EuDASM). The objective was to transfer soil information from paper maps into digital format, with the maximum resolution possible and to preserve the information of paper maps that are vulnerable to deterioration (Panagos et al., 2011). This is one of the most requested sections in ESDAC; it includes many historical maps (starting from the 1920 s) for almost all countries in the World with specific focus in Africa, Asia, and Latin America.

6.3 Social media

Over the past years, ESDAC has increased its presence on social media. In order to reach the broadest possible audience, each release of a new dataset is followed by social media promotion through different platforms (e.g., Twitter, Facebook, and Reddit). Ad-hoc tags (e.g., #ESDAC, #EUSO, #LUCAS18) and storytelling are used to present soil datasets, projects, and related initiatives, with a particular focus on the possible impact in policy development and monitoring. So far, Twitter and Facebook contribute to 1.3% of the total ESDAC annual visits. The enormous potential of social media should be better exploited and a greater integration of social media into ESDAC could contribute to that.

7 FUTURE DEVELOPMENTS

An even stronger ESDAC is among the objectives of the recently established EUSO. In addition, it is inevitable that ESDAC will play an important role in implementing the European Union policy objectives in relation to soils. ESDAC can be an example for thematic portals at national soil institutes. The ingredients for the success of ESDAC are the implementation of an open access data policy, the provision of fully documented datasets (metadata and publications), the presence of an operational and timely helpdesk and keeping ESDAC alive through regular updates, announced to its user-base through informative newsletters.

Thematically, ESDAC is expected to have an increased number of data flows coming from large-scale research projects, monitoring schemes, technological advancements, and policy developments. The European Joint Programme (EJP) for agricultural soils is one of the largest EU-funded research project, running for the period 2020–2024. Among the objectives of the EJP SOIL is to harmonise datasets at continental scale; these new datasets will be hosted in ESDAC (Visser et al., 2020). Another important step forward could be the plethora of soil-related datasets, which are and will be produced in the EU research framework programmes but are not made public in a systematic way. In both cases (EJP SOIL and the HORIZON EUROPE/2020 projects), issues with intellectual property rights are a serious and limiting issue and should be addressed from the outset.

LUCAS 2018 soil survey includes modules for measuring soil biodiversity, pesticides, and microplastics (Orgiazzi et al., 2022). Apart from soil biodiversity, the pesticides module includes data generation in over 3400 locations, mainly from croplands and grasslands, for which 118 different active ingredients and metabolites are measured. These represent the largest pools of soil biological and pollution information available at EU scale, showing great integration potential for indicator development and monitoring of environmental and agricultural policies. Furthermore, data collections, and thus available datasets, are going to increase in coming years. For instance, in LUCAS 2022 soil survey, the number of locations where soil biodiversity data will be measured has been doubled to 2000.

Consideration also needs to be given to accommodate a greater range of products derived from Earth Observation systems and innovative data flows that capture soil condition. Possibilities include proximal sensing systems from precision agriculture and drones to reporting through citizen science initiatives. This will demand an increasing application of big data procedures and the use of artificial intelligence to extract coherent knowledge.

Finally, much of the future new ESDAC datasets could depend on the forthcoming Soil Health Law (2023), as this policy instrument will design the future needs for soil data in the European Union. In addition, the Zero Pollution Action Plan and the Farm to Fork Strategy will require assessments on a range of soil contaminants (e.g., cadmium, arsenic, PFAS, plastics, antibiotics, etc.), nutrient balances (NPK), sediment transport, etc. A specific role will be as a cornerstone of an eventual integrated EU soil monitoring system that builds on data collected through LUCAS and by Member States.

Technically, ESDAC will possibly incorporate additional functionality such as a Web Mapping Application and Services (WMS) to allow users a more direct interaction with ESDAC datasets. A new graphical interface may also make the portal more appealing. Data navigation and search will be based on a web mapping platform where users can use the data according to multiple criteria (space, time and properties). Such developments will be structured according to EU INSPIRE Directive and international interoperability standards (OGC) (Kotsev et al., 2015; Trilles et al., 2017). This will accommodate data sharing and interoperability with Big Earth Observation Data platforms such as Google Earth Engine (GEE), Open Data Cube and others (Gomes et al., 2020; Soille et al., 2018). Another future technical development will be the add-on of an Application Programming Interface (API) suitable for large scale spatial deployments in a standardised way (Schramm et al., 2021). Such an API will care for the immediate availability of large-scale data in a homogenised form and their integration into programming environments for modelling and data analysis.

AUTHOR CONTRIBUTIONS

Panos Panagos: Conceptualization; investigation; writing – original draft; methodology; validation; visualization; writing – review and editing; formal analysis; project administration; data curation; supervision; resources. Marc Van Liedekerke: Conceptualization; supervision; writing – review and editing. Pasquale Borrelli: Writing – original draft; methodology; visualization; writing – review and editing; data curation. Julia Köninger: Methodology; formal analysis; data curation; visualization. Cristiano Ballabio: Conceptualization; investigation; visualization; validation; formal analysis; software; data curation. Alberto Orgiazzi: Conceptualization; formal analysis; supervision; data curation; writing – review and editing; writing – original draft. Emanuele Lugato: Writing – original draft; investigation; writing – review and editing; formal analysis. Leonidas Liakos: Software; formal analysis; data curation; validation; methodology; visualization. Javier Hervas: Writing – review and editing; writing – original draft. Arwyn Jones: Project administration; supervision; funding acquisition. Luca Montanarella: Project administration; supervision; funding acquisition; conceptualization.

DATA AVAILABILITY STATEMENT

All datasets are free to download in ESDAC and they are listed in the web address: https://esdac.jrc.ec.europa.eu/resource-type/datasets.