The validation cases will enable a proof of concept of the EO-approach at different spatial scales, ranging from the regional level (address issues of geographical patterns of interest e.g. crops/ crop residues etc.) to the national (monitoring needed to extrapolate knowledge from the small scale and convey it to the national decision making mechanism).
The criteria for the case study areas are among others: a distribution throughout the bioclimatic regions of Europe with a range of vegetation types, land use and soil types.
The extent of the case study areas will be decided upon together with the members of the National Reporting Centers on soils. The case studies have been chosen in such a way that they coincide with current monitoring and mapping initiatives in the framework of national and international reporting.
The EO-driven approach will be tested in a series of carefully designed validation cases with the active involvement of National Reference Centers of Soil.
The validation cases have been chosen to represent different bioclimatic regions of European territory with a range of vegetation types, land use and soil types:
Southern Belgium, also called Wallonia, covers about 16,900 km2. From the northwest to the southeast, there is an increase in precipitation (from 800 to 1200 mm) along with elevation (from 180 to 690 m) and a decrease of mean annual temperature (from 10 to 8 °C).
In the same direction, a shift occurs from deep sandy loam and silty soils to shallow silt loam and stony soils, from Haplic Luvisol to Distric Cambisol mainly (with Fluvisol on valley bottoms), along with a shift from intensive arable agriculture to more extensive cattle breeding.
Ten agricultural regions are defined on homogeneous characteristics in terms of agricultural practices (types of crops and/or livestock) and crop yields, both directly linked to soil types (e.g., texture, fertility, stoniness, and drainage) and climatic characteristics.
Agricultural regions are now used as references in most of the recent reporting activities linked to agricultural activities and diagnosis of soil quality evolution (such as National Inventory Report –NIR – commanded by EU or the state of the environment report published by the Public Administration of Wallonia).
The first SOC dataset consists of 592 locations from the original National Soil Survey sampling network that were resampled to form the CARBOSOL soil monitoring network dedicated to agricultural soils (croplands and grasslands; 2005-2014;).
The total area of the Czech Republic is 78,865 km2. Slightly more than a half of it (53.7%) is agricultural land, out of which approximately 70% is arable land (38% of the total area of the country), 23% is grassland, the rest are gar-dens, orchards, vineyards, hop fields and other cultures. 33% of the country area is covered by forests. The rest of the country is occupied by urban areas, water bodies etc.
A number of legacy soil datasets including soil organic carbon content were collected in the Czech Republic. These data are available or can be reached under some conditions. The problem of these data is the age and heterogene-ous methodology of sampling design, sampling depth, and soil analyses. Following text presents the basic character-istics of the datasets, general information is summarized in Table 1.
The Systematic soil survey of agricultural land (KPP) was done in 1960-1970. It is the richest source of soil infor-mation so far. In total, 365,615 soil pits were described. In selected soil pits forming approximately one tenth of it (36,942), complete analysis of samples from all horizons was done. Sampling density of the selected pits differed be-tween the types of land-use and between areas with different variation of natural conditions; it ranged between 30 and 180 ha per pit. Localization of the pits is only approximate. Soil monitoring of agricultural land has been carried out regularly by the Central Institute for Supervising and Testing in Agriculture (UKZUZ) since 1992 (Table 2 14). It is done only on selected locations, and it includes a separate set of polluted areas. Agrochemical soil testing (AZP) has been also carried out regularly by the UKZUZ, in this case since 1962; current interval is 6 years. However, SOC determination has been included among the analyses only recently. Forest soil survey (FSS) is done repeatedly since 1996 on all forest soils in the Czech Republic; soil samples from around 200 locations is analyzed every year. Other data are collected also repeatedly since 1980s from 140 selected forest permanent testing areas (TZP); the last sampling in this program was done in 2011. Forest soil pollution survey was done on 120 locations between 2011 and 2012 in frame of a project (KOLEP).
Czech Republic is involved in international program ICP Forest and took part in the BioSoil project surveying forest soils. There are some bilateral projects with neighbouring countries, like INTERREG that mapped soils in the south-western region of the Czech Republic close to the border with Bavaria in Germany. A number of soil data is collected in framework of various research projects aimed on soil pollution, precision agriculture, erosion mapping, etc. The re-sults contain very valuable data. However, they have usually only local or regional extent and quite often they use specific purposive sampling or specific methodology. The availability of these data is often limited, too, as they are produced and owned by different institutions. Some very recent sampling was done on approximately 50 points on agricultural land of the Frydek-Mistek district (NE of the Czech Republic) in 2019, similar number of sampled locations is planned for the year 2020. Very detailed sampling was done on a few agricultural fields in Central Bohemia in 2019. Characteristics of these samples including SOC content are available and can be used as calibration, testing or valida-tion datasets for the WOSOMS project.
In addition, there are some information regarding the source of the data included in the current version of soil data-base and resources for the Czech SOC map provided to Global Soil Partnership (GSP), which may be useful for the de-tails of the study areas (Czech Republic):
- Systematic soil survey – legacy data (from 1960s) (3120 soil pits)
- Interreg project (cross-border, Czech-Germany) (277 soil pits)
- Forest soil pollution study (our project) (120 soil pits)
- Research done by the Department and soil monitoring system (about 1000 soil pits)
The data are collected from the whole territory of the Czech Republic.
Database of the abovementioned data contain profile data, for the purpose of mapping it was rearranged in two groups of data. First set: depth 0 – 30 cm, the second one for depth 30 – 60 cm. For calculation of the organic car-bon stock the bulk density of soil was used. The data were used for preparation of soil map at scale 1:250,000 for both depths. Moreover, the map of both soil organic carbon and humus content expressed as its percentage or its stock were mapped at the scale 1:500,000.
Central Macedonia (Greece)
There is a growing realization amongst policy-makers in Greece that a systemic approach is key to resolving chal-lenges for soil monitoring with regards to national level reporting to UNFCCC and UNCCD, as well as to fulfill the need for a set of EU policies.
The region of interest for the testing and validation of the WOSOMS methodology will be the Region of Central Mace-donia in Greece, around 18,000 km2. At this stage, available soil information originating from different sources have al-ready been identified. This includes completed surveys at national or regional scales (EUSDAC soil map of Greece, Na-tional maps conducted by AUTH>3000 points). Moreover, there are many past projects and research activities, the latter being GEO-CRADLE (an H2020 project), upon whose knowledge the WOSOMS project can build upon to lever-age existing national reference SSLs. The above-mentioned databases contain profile data, for the purpose of map-ping it was rearranged in two groups of data. First set: depth 0 – 30 cm, the second one for depth 30 – 60 cm. For calculation of the organic carbon stock the bulk density of soil was used.
Other assets of the area with regards to access to hyperspectral airborne data (ESA 21018 and EnMAP 2019 cam-paigns) should also be mentioned. The quality of the various datasets (in terms of standardization and harmonization) and availability of metadata have already been evaluated. Detailed field protocols and supporting tools will be used through the in-situ data collection. A toolset will be utilized for (i) recording geo-tagged field observations, (ii) making quality and consistency checks of the observations, and (iii) automated uploading of all data to a central database. Among the selected sampling locations, a maximum of 10% representative sites will be selected to proximal sens-ing tools (e.g. MEMS) for possible future application. In this context, VNIR-SWIR spectra will be acquired in situ to test the accuracy of spectral based soil property predictions in field conditions, to bridge the gap among in situ and spaceborne observations.
A sufficient quantity of soil sample will be taken and will be analyzed (both chemical and spectral analysis are consid-ered) at the Laboratory of Remote Sensing, Spectroscopy and GIS of the Aristotle University in Thessaloniki, Greece. Soil profile description, determination of SOC and classification will be performed in accordance with FAO and GLOSO-LAN standards (AUTH is a member of GLOSOLAN network). The laboratory at AUTH is fully equipped with sensors for spectral analysis (VNIR-SWIR) and their members have proven experience in spectral analysis supported by quality control services. For the spectral analysis, AUTH used standard operating procedures so that the spectra can be comparable to other SSLs (e.g. LUCAS and the Brazilian SSL).
The soil sampling design as well as the analyses aforementioned will draw heavily on the experiences of the AUTH team (responsible to implement the national map for EUSDAC-2015, together with other state-of-the-art digital soil mapping methods and expert knowledge from previous projects (GEO-CRADLE and e-shape).