Impact of data density and endmember definitions on long-term trends in ground cover fractions across European grasslands
Data files
Apr 04, 2025 version files 6.69 GB
-
AR_LND_baseline.tar
979.54 MB
-
AR_LND_S2.tar
951.20 MB
-
AR_LND.tar
4.75 GB
-
README.md
6.58 KB
Abstract
Long-term monitoring of grasslands is pivotal for ensuring continuity of many environmental services and for supporting food security and environmental modelling. Remote sensing provides an irreplaceable source of information for studying changes in grasslands. Specifically, Spectral Mixture Analysis (SMA) allows for quantification of physically meaningful ground cover fractions of grassland ecosystems (i.e., green vegetation, non-photosynthetic vegetation, and soil), which is crucial for our understanding of change processes and their drivers. However, although popular due to straightforward implementation and low computational cost, ‘classical’ SMA relies on a single endmember definition for each targeted ground cover component, thus offering limited suitability and generalization capability for heterogeneous landscapes. Furthermore, the impact of irregular data density on SMA-based long-term trends in grassland ground cover has also not yet been critically addressed.
We conducted a systematic assessment of i) the impact of data density on long‑term trends in ground cover fractions in grasslands; and ii) the effect of endmember definition used in ‘classical’ SMA on pixel- and map-level trends of grassland ground cover fractions. We performed our study for 13 sites across European grasslands and derived the trends based on the Cumulative Endmember Fractions calculated from monthly composites. We compared three different data density scenarios, i.e., 1984-2021 Landsat data record as is, 1984-2021 Landsat data record with the monthly probability of data after 2014 adjusted to the pre‑2014 levels, and the combined 1984-2021 Landsat and 2015-2021 Sentinel-2 datasets. For each site we ran SMA using a selection of site‑specific and generalized endmembers, and compared the pixel- and map-level trends. Our results indicated no significant impact of varying data density on the long-term trends from Cumulative Endmember Fractions in European grasslands. Conversely, the use of different endmember definitions led in some regions to significantly different pixel- and map-level long‑term trends raising questions about the suitability of the ‘classical’ SMA for complex landscapes and large territories. Therefore, we caution against using the ‘classical’ SMA for remote‑sensing‑based applications across broader scales or in heterogenous landscapes, particularly for trend analyses, as the results may lead to erroneous conclusions.
Data description:
Pixel-level trends in green vegetation (gv), non-photosynthetic vegetation (npv), and soil (soil) ground cover fractions derived and analyzed in the course of a paper:
Lewińska K.E., Okujeni A., Kowalski K., Lehnamm F., Radeloff V.C., Leser U., Hostert P., (2025) Impact of data density and endmember definitions on long-term trends in ground cover fractions across European grasslands, Remote Sensing of Environment, 323, https://doi.org/10.1016/j.rse.2025.114736
The datasets are generated as GTifs and distributed as three .tar archives:
AR_LND: comprising AR(1) trends derived based on the 1984-2021 Landsat (TM, ETM+, and OLI) time seriesAR_LND_baseline: comprising AR(1) trends derived based on 1984-2021 Landsat (TM, ETM+, and OLI) with the pixel-level probability of cloud-, snow-, and shade-free observation for the 2015-2021 period adjusted to the 1984-2014 baseline.AR_LND_S2: comprising AR(1) trends derived based on combined 1984-2021 Landsat (TM, ETM+, and OLI) and 2015-2021 Sentinel 2 (MSI) time series
Each archive comprises several directories following the SIT_END naming convention, with SIT representing the two or three letter code of a test site and END representing the endmember set used in the unmixing process.
The SIT codes are: ALP - Apline; BX - Belgium; CR - Crete; DE - Germany; ES - Spain; FR - France; IE - Ireland; LX - Luxembourg; PL - Poland; RO - Romania; SA - Sardinia; SE - Sweden; UK - United Kingdom.
The endmember END sets area: ATL - ; SA - Mediterranean ; SE - Boreal; TEM - Continental. When the END code is the same as the SIT code this signifies the ‘local’ set of endmembers i.e. collected on the specific site. Other codes represent regional ets of endmembers i.e., generalized over a specific biogeographical region. for SE and SA sites the local and regional endmembers are identical.
Each SIT_END folder further comprises between one and four subfolder following the Xxxxx_Yyyyy naming convention, where xxxx and yyyy give, correspondingly, X and Y coordinates of a tile in a FORCE data cube. We used the standard definition for the data cube coordinates given in: https://force-eo.readthedocs.io/en/latest/howto/datacube.html
Finally, each Xxxxx_Yyyyy tile comprises three GTiffs with AR(1) trends in green vegetation (gv), non-photosynthetic vegetation (npv), and soil (soil) ground cover fractions, gv_AR.tif, npv_AR.tif, and soil_AR.tif, respectively. Each *_AR.tif comprises 4 bands:
band 1: AR(1) trend in respective ground cover fraction based on 1984-2021 time series of annual Cumulative Endmember Fractions
band 2: intercept of the AR(1) trend in band 1
band 3: pixel-level P-value of the AR(1) trend in band 1
band 4: strength of temporal autocorrelation in the 1984-2021 time series.
All data are available at 30-m resolution and in the LAEA projection defined as:
gdalsrsinfo -v 'PROJCS["ETRS89 / LAEA Europe",GEOGCS["ETRS89",DATUM["European_Terrestrial_Reference_System_1989",SPHEROID["GRS 1980",6378137,298.257222101,AUTHORITY["EPSG","7019"]],TOWGS84[0,0,0,0,0,0,0],AUTHORITY["EPSG","6258"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AUTHORITY["EPSG","4258"]],PROJECTION["Lambert_Azimuthal_Equal_Area"],PARAMETER["latitude_of_center",52],PARAMETER["longitude_of_center",10],PARAMETER["false_easting",4321000],PARAMETER["false_northing",3210000],UNIT["metre",1,AUTHORITY["EPSG","9001"]],AUTHORITY["EPSG","3035"]]'
Validate Succeeds
PROJ.4 : '+proj=laea +lat_0=52 +lon_0=10 +x_0=4321000 +y_0=3210000 +ellps=GRS80 +towgs84=0,0,0,0,0,0,0 +units=m +no_defs '
OGC WKT :
PROJCS["ETRS89 / LAEA Europe",
GEOGCS["ETRS89",
DATUM["European_Terrestrial_Reference_System_1989",
SPHEROID["GRS 1980",6378137,298.257222101,
AUTHORITY["EPSG","7019"]],
TOWGS84[0,0,0,0,0,0,0],
AUTHORITY["EPSG","6258"]],
PRIMEM["Greenwich",0,
AUTHORITY["EPSG","8901"]],
UNIT["degree",0.0174532925199433,
AUTHORITY["EPSG","9122"]],
AUTHORITY["EPSG","4258"]],
PROJECTION["Lambert_Azimuthal_Equal_Area"],
PARAMETER["latitude_of_center",52],
PARAMETER["longitude_of_center",10],
PARAMETER["false_easting",4321000],
PARAMETER["false_northing",3210000],
UNIT["metre",1,
AUTHORITY["EPSG","9001"]],
AUTHORITY["EPSG","3035"]]
Methodology
For detailed description of the methodology used in pre-processing and processing of the shared data, please see the linked publication: Lewińska K.E., Okujeni A., Kowalski K., Lehnamm F., Radeloff V.C., Leser U., Hostert P., (2025) Impact of data density and endmember definitions on long-term trends in ground cover fractions across European grasslands, Remote Sensing of Environment, 323, https://doi.org/10.1016/j.rse.2025.114736
Sharing & Access information
CC license: CC 00
Data are available at datadryad.org as: 10.5061/dryad.fqz612k3r
The data were derived based on freely available Landsat surface reflectance Level 2, Tier 1 (Collection 2) scenes from 1984 through 2021 (Thematic Mapper (TM): doi.org/10.5066/P918ROHC;, Enhanced Thematic Mapper (ETM+): https://doi.org/10.5066/P9TU80IG, and Operational Land Imager (OLI): https://doi.org/10.5066/P975CC9B) and Sentinel-2 TOA reflectance Level-1C (pre Collection-1: https://doi.org/10.5270/S2_-d8we2fl). All used satellite data are in the public domain thus adhere to the CC0 licence.
Disclaimer:
The authors accept no responsibility for errors or omissions in this work and shall not be liable for any damage caused by these.
Funding Information
We gratefully acknowledge support from the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – Project-ID 414984028 – SFB 1404 FONDA.
Please cite the data as:
Lewińska K.E., Okujeni A., Kowalski K., Lehnamm F., Radeloff V.C., Leser U., Hostert P., Impact of data density and endmember definitions on long-term trends in ground cover fractions across European grasslands., Dryad dataset, 10.5061/dryad.fqz612k3r
and
Lewińska K.E., Okujeni A., Kowalski K., Lehnamm F., Radeloff V.C., Leser U., Hostert P., (2025) Impact of data density and endmember definitions on long-term trends in ground cover fractions across European grasslands, Remote Sensing of Environment, 323, https://doi.org/10.1016/j.rse.2025.114736
For the detailed description of the methodology used to derived the datasets please refer to the related publication:
Lewińska K.E., Okujeni A., Kowalski K., Lehnamm F., Radeloff V.C., Leser U., Hostert P., "Impact of data density and endmember definitions on long-term trends in ground cover fractions across European grasslands". under review in Remote Sensing of Environment, Available as preprint: https://doi.org/10.31223/X5D43N