Аннотация:Three indicators were recommended by the UNCCD to follow the progress in the
monitoring progress towards achieving the Land Degradation Neutrality targets at the national
level: land cover change (LCC), land productivity dynamics (LPD) and trends in soil organic
carbon (SOC). All of them are based on open source datasets with global coverage: ESA CCI
Land Cover product time series from 1992-2015, JRC’s LPD Dataset and SoilGrids (ISRIC).
The use of these datasets are supposed to ensure comparable and standardized results in terms
of revealing trends in land degradation for different evaluation periods based on “one-out – allout
rule” and to support spatially-explicit framework for land monitoring and revealing
hotspots. In the meantime, at the national level all the initial data and their assessments results
need for trusted refinements, accuracy assessment and validation based on existing comparable
data sources. In the case of Russia with its huge territory elaborative approach to assessment
and interpretation trends in LDN indicators is necessary to keep in mind high and spatially
heterogeneous diversity of ecosystems, climate types and land use patterns.
In the absence of national time-consistent land cover products in Russia ESA CCI Land
cover products for the period 2000-2015 was used without default in LDN methodology
aggregation of land cover classes. Such approach affords to interpret numerous land cover
transitions types separately for the regions with different landscape conditions and drivers of
land degradation, more accurately and precisely reveal and evaluate the land degradation and
land restoration spatial patterns in the semiarid areas. The accuracy assessment of cumulative
LCC from ESA CCI products founds the overestimation of cropland expansion in steppe zone
and underestimation of boreal forest loss for the Russian territory. The results of LPD
assessment are consistent with trends revealed from different national evaluations and
inventory, and also well verified with changes in crop yield and forest stocks. The main trouble
presents SOC assessments derived from SoilGrids as this product have low accuracy and
contain the great discrepancies both in absolute values as in their spatial distribution.