National reporting has become a central element in steering the sustainability transformation in response to global climate and land use change. The Copernicus and the Landsat programs are hence cornerstones for quantifying SDG indicators and metrics at national level. In Europe, the Common Agricultural Policy (CAP) and quantifications of agricultural ecosystem services related to productivity, groundwater and soil quality are another field where continuous Earth Observation (EO) data are of great importance. Given the urgency of climate change impact mitigation, national carbon accounting and biodiversity monitoring are also core action fields for Copernicus- and Landsat-based mapping and monitoring. The free and open data policy and improvements towards the provision of Analysis Ready Data (ARD) at ESA/EU on the one hand and USGS/NASA on the other hand have accordingly triggered major improvements in national monitoring activities that were out of reach only a decade ago.
We here demonstrate the unprecedented mapping and monitoring capabilities of combined Sentinel-2, Sentinel-1, and Landsat-based national monitoring across different domains, namely croplands, grasslands, and urban / built-up systems. The agricultural domain and urban systems are the major sources of greenhouse gas emissions both related to land use and related to processes along the value chains they build upon. Regarding agricultural ecosystems, land use intensity can be used as a proxy for measuring the pressure on ecosystems in general and greenhouse gas emission potential in particular. Information on grassland use intensity related to mowing events, grazing impacts or both provide us with insights on biomass extraction that directly relate to subsequent carbon emissions. In cropping systems, monitoring of crop rotation patterns allows estimating land use intensity. Finally, while settlements make up for only a minor fraction of land cover, resource use due to construction, maintenance and operation of material stocks directly links to carbon emitted across the value chain when building materials are extracted, processed and traded, or when housing and traffic infrastructures are built and utilized.
Blickensdörfer et al. (in review, https://ows.geo.hu-berlin.de/webviewer/croptypes/) created the first multi-year, wall-to-wall and high-resolution crop type map of Germany that allows for insights into the crop rotation systems across the country. Our approach proved that mapping 24 agricultural land-cover classes was possible for 2017 to 2019, even though the meteorological conditions strongly differed between the three years – including severe drought. Integrating all available Sentinel-2, Landsat and Sentinel-1 observations with environmental data demonstrated the feasibility of providing crop type mapping even under extremely varying boundary conditions in between the years. We are now for the first time able to bring the crop rotation spatial pattern in alignment with soil and weather patterns for risk and vulnerability analyses, providing substantially improved suggestions for climate change adaptation.
Schwieder et al. (2021) recently showed that mowing events in grasslands across entire Germany can be identified based on residuals from an assumed undisturbed phenology using combined Sentinel-2 / Landsat-8 time series. We mapped mowing events for 2017 to 2020 for all permanent grasslands in Germany as classified by Blickensdörfer et al. (in review). We found that the detection of mowing events was less influenced by data availability when at least 16 cloud-free observations were available in the grassland season, while the distribution of clear sky observations throughout the season is critical. Hotspots of highly intensively managed grasslands were identified in the alpine foreland in Southern Germany as well as in the lowlands in the Northwest of Germany. We could also prove a tendency to lower management intensity in the extremely dry year 2018. As mowing events are a direct indicator of grassland use intensity, this is also an important step towards moving from mere land cover-related observations to land use-focused analyses.
Mapping urban and other built-up areas is a completely novel field for remote sensing centered national accounting. High-resolution mapping of material stocks of buildings and infrastructure, based on Sentinel-1, Sentinel-2, and Open Street Map (OSM) data has been developed only recently. Haberl et al. (2021; https://ows.geo.hu-berlin.de/webviewer/stocks/) for the first time estimated the total mass of buildings and infrastructures for Germany and Austria in a high-resolution and spatially explicit manner, drawing on interdisciplinary collaboration with industrial ecology and detailed research on building typologies and construction standards. Material stocks amounted to ca. 5 Gt in Austria and ∼38 Gt in Germany for 2018, which makes up for 540 t/cap of average building material usage in Austria and ca. 450 t/cap in Germany. These findings were only possible due to innovative ways to estimate national land cover fractions from Sentinel-2 (Schug et al. 2020; https://ows.geo.hu-berlin.de/webviewer/land-cover-fractions/) and national building heights from Sentinel-1 and Sentinel-2 data (Frantz et al. 2020; https://ows.geo.hu-berlin.de/webviewer/building-height/).
All these findings specifically draw on the opportunities related to the improved temporal resolution offered by the Sentinels, Landsat data integration and heritage, along with well-calibrated spectral bands and improved spatial resolution. Moreover, the remote sensing community’s option space has greatly improved with readily pre-processed data and opportunities to integrate information from multi-sensor missions and additional geodata. Summarizing, given the momentum gained in the last years, we encourage further improvements in ARD provision, as well as further efforts to integrate third-party missions and additional geodata in a free and open fashion. Along those lines, providing non-commercial cloud processing opportunities for science is of utmost importance to ensure steady innovation bridging between basic science and application.
Blickensdörfer, L., Schwieder, M., Pflugmacher, D., Nendel, C., Erasmi, S., & Hostert, P. (accepted). Multi-year national-scale crop type mapping with combined time series of Sentinel-1, Sentinel-2 and Landsat 8 data. Remote Sensing of Environment.
Frantz, D., Schug, F., Okujeni, A., Navacchi, C., Wagner, W., van der Linden, S., & Hostert, P. (2021). National-scale mapping of building height using Sentinel-1 and Sentinel-2 time series. Remote Sensing of Environment, 252, 112128. https://doi.org/10.1016/j.rse.2020.112128
Haberl, H., Wiedenhofer, D., Schug, F., Frantz, D., Virág, D., Plutzar, C., ... & Hostert, P. (2021). High-resolution maps of material stocks in buildings and infrastructures in Austria and Germany. Environmental science & technology, 55(5), 3368-3379. https://doi.org/10.1021/acs.est.0c05642
Schug, F., Frantz, D., Okujeni, A., van Der Linden, S., & Hostert, P. (2020). Mapping urban-rural gradients of settlements and vegetation at national scale using Sentinel-2 spectral-temporal metrics and regression-based unmixing with synthetic training data. Remote Sensing of Environment, 246, 111810. https://doi.org/10.1016/j.rse.2020.111810
Schwieder, M., Wesemeyer, M., Frantz, D., Pfoch, K., Erasmi, S., Pickert, J., Nendel, C., & Hostert, P. (2021). Mapping grassland mowing events across Germany based on combined Sentinel-2 and Landsat 8 time series. Remote Sensing of Environment, 112795. https://doi.org/10.1016/J.RSE.2021.112795