The Global Multi-resolution Terrain Elevation Data 2010 (GMTED2010) provides a new level of detail in global topographic data. Previously, the best available global DEM was GTOPO30 with a horizontal grid spacing of 30 arc-seconds. The GMTED2010 product suite contains seven new raster elevation products for each of the 30-, 15-, and 7.5-arc-second spatial resolutions and incorporates the current best available global elevation data. The new elevation products have been produced using the following aggregation methods: minimum elevation, maximum elevation, mean elevation, median elevation, standard deviation of elevation, systematic subsample, and breakline emphasis. Metadata have also been produced to identify the source and attributes of all the input elevation data used to derive the output products. Many of these products will be suitable for various regional continental-scale land cover mapping, extraction of drainage features for hydrologic modeling, and geometric and radiometric correction of medium and coarse resolution satellite image data. The global aggregated vertical accuracy of GMTED2010 can be summarized in terms of the resolution and RMSE of the products with respect to a global set of control points (estimated global accuracy of 6 m RMSE) provided by the National Geospatial-Intelligence Agency (NGA). At 30 arc-seconds, the GMTED2010 RMSE range is between 25 and 42 meters; at 15 arc-seconds, the RMSE range is between 29 and 32 meters; and at 7.5 arc-seconds, the RMSE range is between 26 and 30 meters. GMTED2010 is a major improvement in consistency and vertical accuracy over GTOPO30, which has a 66 m RMSE globally compared to the same NGA control points. In areas where new sources of higher resolution data were available, the GMTED2010 products are substantially better than the aggregated global statistics; however, large areas still exist, particularly above 60 degrees North latitude, that lack good elevation data. As new data become available, especially in areas that have poor coverage in the current model, it is hoped that new versions of GMTED2010 might be generated and thus gradually improve the global model.

Connectivity assessments and corridor delineation are key contributions to landscape management and biodiversity conservation. We examined the influence of three potentially crucial factors on the results of connectivity analyses, using the two subpopulations of the endangered brown bear in the Cantabrian Range (NW Spain) as a case study. First, we evaluated the spatial resolution of vegetation data, using three types of datasets ranging from coarse resolution land-cover maps to high-resolution LiDAR data. Second, the seasonal variation in the distribution of habitat resources and in the species use of the landscape. Third, multi-annual periods with different population status. The estimates of subpopulation isolation (effective distances) and the trajectory of the identified corridors were substantially influenced by (i) the spatial resolution of vegetation data; the more robust results were obtained when incorporating fine-scale LiDAR data; (ii) the season over which species occurrence data and landscape characteristics were considered; the spring mating season yielded higher connectivity estimates than any other season; (iii) the status of the populations, with higher landscape connectivity estimated for expanding populations. Our study reveals that the use of coarse-resolution data may underestimate the resistance of the non-habitat landscape matrix to species movements. The use of year-round estimates of habitat connectivity may miss the key seasonal temporal windows for species movements. Landscape resistance may be overestimated when data from periods with declining or restricted populations are used. We recommend carefully disentangling the effects of demography and landscape heterogeneity on realized species dispersal movements for improving the insights from connectivity modelling.

*default setting.

This dataset provides high-resolution Leaf Area Index (LAI) reference maps for validating global LAI products. It includes a total of 258 high-resolution reference LAI rasters which are available across 105 locations in 25 countries, with spatial resolutions depending on the sensor (e.g., 30m for Landsat, 10m for SPOT). The maps contain LAI effective values and are organized by validation study, country, and location, with standardized filenames for easy processing. The dataset is available for public access on Zenodo and can be used for further analysis and validation of LAI products. Five validation studies were included and the respective articles can be found below.

Validation Studies;

• ImagineS: Fuster, B., S´anchez-Zapero, J., Camacho, F., García-Santos, V., Verger, A., Lacaze, R., Weiss, M., Baret, F., Smets, B., 2020. Quality assessment of PROBA-V LAI, fAPAR and fCOVER collection 300 m products of Copernicus global land service. Remote Sens.12 (6), 1017.
• Valeri: Baret, F., Weiss, M., Allard, D., Garrigues, S., Leroy, M., Jeanjean, H., Fernandes, R., Myneni, R., Privette, J., Morisette, J., 2021. VALERI: a network of sites and a methodology for the validation of medium spatial resolution land satellite products ⟨hal-03221068⟩.
• Bigfoot: Gower, S.T., Kirschbaum, A.A., 2008. BigFoot Field Data for North American Sites, 1999–2003. ORNL Distributed Active Archive Center.
• ValLAI_Crop: Song, B., Liu, L., Du, S. et al. ValLAI_Crop, a validation dataset for coarse-resolution satellite LAI products over Chinese cropland. Sci Data 8, 243 (2021).
• CNERN: Fang, Hongliang; Zhang, Yinghui; Wei, Shanshan; Li, Wenjuan; Ye, Yongchang; Sun, Tao; Liu, Weiwei (2019): The field measurements and high resolution reference LAI data in Hailun and Honghe, China [dataset]. PANGAEA

Dataset description;

• LAI Reference Dataset.zip: Contains 258 high resolution standardized LAI rasters organized into subfolders (i.e. country, validation project, year, location )
• Moderate Resolution LAI Products.zip: Contains moderate resolution LAI raster products (MODIS, PROBA-V, GLASS) and MODIS NDVI raster for 9 reference sites. These products were used to validate the high resolution LAI rasters that exist in the dataset.
• LAI_Reference_Dataset_Locations_v2.xlsx: A list of high resolution LAI raster locations and their point coordinates with important details
• Readme_LAI_Dataset.txt: Summary of essential information

Training and test AUC values (mean) for both models and measured model performance for the high-resolution (1 m) and low-resolution model (50 m).

Landsat 30m resolution observations provide sufficient spatial details for monitoring land surface and changes. However, the 16-day revisit cycle and cloud contamination have limited its use for studying global biophysical processes, which evolve rapidly during the growing season. Meanwhile, MODIS sensors aboard the NASA EOS Terra and Aqua satellites provide daily global observations valuable for capturing rapid surface changes. However, the spatial resolution of 250m to 1000m may not good enough for heterogeneous areas. To better utilize Landsat and MODIS data, the Spatial and Temporal Adaptive Reflectance Fusion Model (STARFM) was developed (Gao et al., 2006). The STARFM algorithm uses spatial information from fine-resolution Landsat imagery and temporal information from coarse-resolution MODIS imagery to produce estimates of surface reflectance that are high resolution in both space and time. In essence, the collection of daily MODIS imagery and seasonal Landsat imagery allows the generation of synthetic daily Landsat-like views of the Earth’s surface. The STARFM algorithm uses comparisons of one or more pairs of observed Landsat/MODIS maps, collected on the same day, to predict maps at Landsat-scale on other MODIS observation dates. STARFM was initially developed at the NASA Goddard Space Flight Center by Dr. Feng Gao. This version (v1.2) has been greatly improved in computing efficiency (e.g. one run for multiple dates and parallel computing) for large-area processing (Gao et al., 2015). Additional improvements (e.g. Landsat and MODIS images co-registration, daily MODIS nadir BRDF-adjusted reflectance) in the operational data fusion system (Wang et al., 2014) are beyond the STARFM program and are not included in this package. Improvement and continuous maintenance are being undertaken in the USDA-ARS Hydrology and Remote Sensing Laboratory (HRSL), Beltsville, MD by Dr. Feng Gao. Resources in this dataset:Resource Title: STARFM. File Name: Web Page, url: https://www.ars.usda.gov/research/software/download/?softwareid=432&modecode=80-42-05-10 download page

Coarse-grid weather and climate models rely particularly on parameterizations of cloud fields, and coarse-grained cloud fields from a fine-grid reference model are a natural target for a machine-learned parameterization. We machine-learn the coarsened-fine cloud properties as a function of coarse-grid model state in each grid cell of NOAA's FV3GFS global atmosphere model with 200 km grid spacing, trained using a 3-km fine-grid reference simulation with a modified version of FV3GFS. The ML outputs are coarsened-fine fractional cloud cover and liquid and ice cloud condensate mixing ratios, and the inputs are coarse model temperature, pressure, relative humidity, and ice cloud condensate. The predicted fields are skillful and unbiased, but somewhat under-dispersed, resulting in too many partially-cloudy model columns. When the predicted fields are applied diagnostically (offline) in FV3GFS's radiation scheme, they lead to small biases in global-mean top-of-atmosphere (TOA) and surface radiative fluxes. An unbiased global-mean TOA net radiative flux is obtained by setting to zero any predicted cloud with grid-cell mean cloud fraction less than a threshold of 6.5%; this does not significantly degrade the ML prediction of cloud properties. The diagnostic, ML-derived radiative fluxes are far more accurate than those obtained with the existing cloud parameterization in the nudged coarse-grid model, as they leverage the accuracy of the fine-grid reference simulation's cloud properties.This dataset provides the coarsened fine-grid model outputs needed to run the nudged coarse climate model, including running with prescribed coarsened fine-grid cloud fields and to train the ML model that predicts coarsened-fine cloud fields as functions of nudged coarse model state.

The dataset is High-resolution climate change projections for Queensland. The high-resolution (10 km spatial grid spacing) has been generated using global variable resolution CSIRO CCAM-CABLE model by the Department of Science, Information Technology and Innovation (DSITI) in collaboration with CSIRO Marine and Atmospheric Research. This high-resolution dynamically downscaled data is an extension of CSIRO-BoM Climate Change in Australia: Projections for Australia's NRM Regions (http://www.climatechangeinaustralia.gov.au/en/climate-projections/ ) which were based on the data from coarse resolution global climate projections completed under the CMIP5 (Coupled Model Intercomparison Project Phase 5) to support the development of the 5th Assessment Report of IPCC released during 2013-2014. The high resolution projections were developed in two stages. In first time varying sea surface temperature and sea ice from selected global climate models (CMIP5) were bias corrected and used to force the 50 km global uniform resolution CCAM-CABLE model for the period 1950-2099 under RCP8.5 emission pathway. In stage two the global stretched version of CCAM-CABLE model at 10 km spatial resolution over Queensland region (9.5-32oS – 132-158oE) was used. Simulations for period 1980-2009 were completed using spectral nudging at 6-hours interval from 50 km simulations (stage 1). Both stages of dynamical downscaling follow the Coordinated Regional Downscaling Experiment (CORDEX - http://www.cordex.org/ ) experimental protocol. The aim CORDEX is to use a coordinated approach to bridge the gap between the climate modelling community and end users of climate information across the globe by providing high-resolution climate change projections for regional scale adaptation. The following CMIP5 global climate model simulations of historical and future climate were downscaled in stage one and two: Australian CSIRO-BoM ACCESS1.0 and ACCESS1.3, US NCAR CCSM4, French CNRM-CM5, US NOAA GFDL-CM3, UK HadGEM2, Germany MPI-ESM-LR and Norway NorESM1-M. There is in total 9 models (8 distinct climate models and ensemble average). There are around 20 different variables for each model and 17 different derived changes (different 20 years long averaging periods into future eg. 2020, 2030..2090 per variable) In total around 3060 files with total size at this stage <10 Gb. This data represents projected changes from historical period eg. changes at 2050 represent difference between 2040-2059 – 1986-2005. In second stage we will release daily (frequency of data) data ~ 720Gb size and hourly data for selected variables.

Mean of hit, false alarm (FA), d’ and VWM capacity estimate (K) in all conditions. (SC: simple shape change; CC: cross-category change; WC: within-category change). (DOC)

The LULC HRL Map is produced from a combination of multi-sources data: the French national topographic database; the Land Parcel Identification System (LPIS) database; and Corine Land Cover. The LULC HRL classification contains 11 land cover categories: 11 Industrial or Commercial buildings and other Facilities 12 Agricultural buildings 13 Low-rise Residential or Mixed buildings 14 High-rise Residential or Mixed buildings 2 Fields 3 Meadows/Grassy plots 4 Bushes/Shrubs 5 Trees/Forest 6 Vineyards 7 Water bodies 8 Others artificial surfaces The LULC HRL Map is produced from a combination of multi-sources data: the French national topographic database (Institut national de l’information géographique et forestière, the French national geographic institute); the Land Parcel Identification System (LPIS) database (Agency for Services and Payment, French public institution responsible for the implementation of national and European public policies; Integrated Administration and Control System, European Union); and Corine Land Cover (European Environment Agency, Joint Research Center, European Union). The topographic database contains a land cover description employed for topographic map production at a scale of 1:25 000, with a minimum unit of collection of approximately 8 ha. The information is relatively precise on the contours of urban areas (buildings), road and rail infrastructures, hydrography, and trees and shrubs; however, it does not make it possible to distinguish the land uses within the agricultural, forested, or natural areas. The LPIS database, which draws on the digital cadastral database (1:500–1:5000), allows us to identify those agricultural areas for which subsidies are sought under the European Common Agricultural Policy (CAP). It was used to determine the agricultural land-use (grass-like vegetation and arable land) on the scale of cadastral parcels. Corine Land Cover (CLC) is thematically much richer, in particular in agri- cultural areas, but its spatial resolution, which is rather coarse (approximately 1:100 000), means it cannot identify the nature of a polygon of less than 25 ha. Despite its rather coarse resolution, CLC has a thematically richer land-use nomenclature than can be used to refine plant cover. The land-cover information layer was constructed in two steps. The first was to generate a simplified geometry of land use in vector form (polygons and lines). The operation begins by detecting the “polygonal skeleton” that integrates roads, railways, and the hydrographic network attributing to them a footprint proportional to their width. Next are added (1) agricultural surface features from the LPIS (field, meadow, orchard, other agriculture use); (2) plant-covered areas, mostly forest and orchard; and (3) artificialized surfaces (buildings, quarries, parking areas, etc.). Each addition is made by masking and expansion so as to approximate the “polygonal skeleton”. The features not described in the topographic database and the LPIS are categorized as “unidentified polygons”. Some of this class is marked down as grassland-lawn using CLC classes “321” (Natural grasslands) and “231” (Pastures). Processing is done with the PostGIS functionalities: intersection, union, dilation, erosion, etc. of polygons or lines (PostGIS, 2018). This stage enables eight land-use categories to be defined: (1) urban footprints, (2) fields, (3) meadows, (4) forests, (5) orchards, (6) rivers and water bodies, (7) road and rail infrastructure footprints, (8) unidentified polygons. This first vectorial geometric model is changed into a 5m resolution raster layer and then supplemented to produce a land-use layer com- patible with the landscape analysis contemplated. Categories (4), (6) and (7) describing relatively homogeneous and straightforward landscape features were kept unchanged. The improvement described below was primarily for heterogeneous and complex landscape features (categories (1), (2), (3) and (5)) that are replaced by simple landscape objects (buildings, mineral surfaces, copses, fields, grass-covered areas, etc.). The improvement also covers pixels in category (8). Pixels of the urban footprint (1) are differentiated into three types of landscape items: the built area, parking areas, and urban plant cover. The built area is incrusted by distinguishing its height and function: (11, LRM) Low-rise Residential or Mixed buildings (< 12m∼1–2 storeys); (12, HRM) High-rise Residential or Mixed buildings (≥12m∼3 storeys and more); (13, ICF) Industrial or Commercial buildings and other Facilities; (14) agricultural buildings. Parking areas were also created around some buildings and classified as category (7): a 5m (1 pixel) buffer around HRM polygons and ICF polygons between 50 and 999m2; a 25m buffer for ICF polygons of 1000m2 (5 pixels) and more. The buffer sizes were established from existing planning and building codes. Non-built and non-parking areas in the urban footprint are converted into plant cover in the following proportions: grass 50% of pixels; Trees 25%; shrubs and bushes 25%. These proportions are based on the visual identification and quantification of green areas/ expanses in built the environment using orthophoto images. This is done by first converting non-built and non-parking areas into grass pixels and then drawing tree pixels and shrub and bush pixels at random. For the field (2) and meadow (3) categories identified with tree cover (presence of trees in CLC), 10% of randomly drawn pixels are converted into trees. The pixels classified as orchards (5) and that are within a polygon classified as vineyard (221) in CLC are reclassified as vineyard. The remaining pixels are first converted into grass and then into shrubs and bushes by randomly drawing 70% of the pixels. Pixels in category (8), “unidentified polygons”, are reclassified by comparison with the CLC polygons.

While modelling habitat suitability and species distribution, ecologists must deal with issues related to the spatial resolution of species occurrence and environmental data. Indeed, given that the spatial resolution of species and environmental datasets range from centimeters to hundreds of kilometers, it underlines the importance of choosing the optimal combination of resolutions to achieve the highest possible modelling prediction accuracy. We evaluated how the spatial resolution of land cover/waterbody datasets (meters to 1 km) affect waterbird habitat suitability models based on atlas data (grid cell of 12×11 km). We hypothesized that the area, perimeter and number of waterbodies computed from high resolution datasets would explain distributions of waterbirds better because coarse resolution datasets omit small waterbodies affecting species occurrence. Specifically, we investigated which spatial resolution of waterbodies better explain the distribution of seven waterbirds nesting on ponds/lakes with areas ranging from 0.1 ha to hundreds of hectares. Our results show that the area and perimeter of waterbodies derived from high resolution datasets (raster data with 30 m resolution, vector data corresponding with map scale 1:10,000) explain the distribution of the waterbirds better than those calculated using less accurate datasets despite the coarse grain of the species data. Taking into account the spatial extent (global vs regional) of the datasets, we found the Global Inland Waterbody Dataset to be the most suitable for modelling distribution of waterbirds. In general, we recommend using land cover data of a resolution sufficient to capture the smallest patches of the habitat suitable for a given species’ presence for both fine and coarse grain habitat suitability and distribution modelling.

This repository contains the processed data and plotting routines associated with the article Holmes, Groeskamp, Stewart and McDougall (2022), Sensitivity of a Coarse-Resolution Global Ocean Model to a Spatially Variable Neutral Diffusivity, Journal of Advances in Modeling Earth Systems (JAMES), doi: 10.1029/2021MS002914, http://dx.doi.org/10.1029/2021MS002914 The contents includes post-processed data output from the 1-degree ACCESS-OM2 ocean-sea-ice model simulations and the python/jupyter plotting routines required to make the plots. The processing script is Holmes2022JAMES_Neutral_Diffusion_ACCESS-OM2_Plotting_Script.ipynb. The data files consist of time-averages or time series of certain metrics processed using NCO tools from the raw ACCESS-OM2 simulation output.

To improve the spatial resolution and preserve the spatial fidelity of TROPOMI Solar-Induced chlorophyll Fluorescence (SIF), we developed a global 0.005° SIF product (RSIF) covering May 2018 to December 2020 using the proposed One-Step Learned Spatial Redistribution Convolutional Neural Network (OSRNet). OSRNet directly redistributes coarse-resolution TROPOMI SIF into fine grids using high-resolution auxiliary variables, including MODIS reflectance, ERA5 reanalysis, GEBCO DEM, and cos(SZA). Validation against both original TROPOMI SIF and long-term tower-based SIF from five flux sites shows that RSIF improves agreement with coarse-resolution inputs and better captures fine-scale spatial details compared to traditional downscaling methods.

This record serves as the parent dataset for the RSIF product covering the years 2018, 2019, and 2020. The complete dataset is split into three annual subsets due to Zenodo’s file size limitations:

• 2018: doi:10.5281/zenodo.16787900

• 2019: doi:10.5281/zenodo.16788564

• 2020: doi:10.5281/zenodo.16788879

Users are encouraged to cite this parent record when referencing the complete dataset, and to cite individual year records when using specific annual subsets.

In the high northern latitudes the vegetation cover and land surface structure is affected by seasonal freeze/thaw dynamics of the uppermost permafrost layer (active layer). The land cover in the arctic regions is characterized by low vegetation species (shrubs, grasses, mosses) in the northernmost regions as well as the boreal forest in the southern parts.The pan-arctic data product presented here is based on user requirements which were defined in the ESA DUE Permafrost for the coarse resolution vegetation observation variable. The user requirements have shown the need of percentage cover information for different vegetation physiognomy and barren areas. Therefore, fractional cover information for trees, shrubs, herbaceous and non-vegetated areas were extracted using coarse resolution global land cover products in a harmonization approach (Urban et al. 2010, doi:10.1127/1432-8364/2010/0056).As input MODIS Land Cover, GlobCover, Synmap as well as MODIS VCF (Vegetation Continuous Fields) were used. In the harmonization approach percentage cover values for different vegetation physiognomy and barren areas were extracted using the class description of each land cover legend. Therefore it was feasible to convert thematic classes to fraction cover values for each pixel. The fractional vegetation cover product has a spatial resolution of 1 km. As this dataset provides percentage information instead of thematic classes, a reduction of the spatial resolution for the integration in modeling approaches (spatial resolution 25 km or 0.5°) is feasible (Urban et al. 2010, doi:10.1127/1432-8364/2010/0056).In the version 1 product we found, that there are cover information from trees north of the tree line (Circumpolar Arctic Vegetation Map, CAVM - Walker et al. 2005), which are reasoned by misclassification of the forest classes in the global land cover datasets. Based on these findings, the CAVM, which provides spatial information about the tree line, were used as thematic information to mask incorrect percentage tree cover information in the tundra regions. Hence, the product version 2 is providing percentage cover information for shrubs, herbaceous and non-vegetated areas in the high latitude tundra regions. No modifications have been done for the boreal regions in the version 2 product.A validation of the version 2 product was done using Landsat imagery with a resolution of 30 m. To assess the accuracy, the Landsat imagery was classified into the four components of the version 2 products (trees, shrubs, herbaceous and non-vegetated areas). By using chessboards/fishnet segments with a resolution of 1 km, the percentage cover of the classes from the Landsat classification are compared with the components of the pan-arctic fractional cover product. The validation results has shown very good agreement for the taiga and tundra areas (trees 83 - 97 %, shrubs 87 - 91 %, herbaceous 66 - 78 % and non-vegetated areas 74 - 85 %). Within the taiga-tundra transitions zone, lower accuracies have been found. This is reasonable since these regions are characterized by heterogeneous landscapes, which are hardly to classify with coarse resolution earth observation data.See "other version" for version 1 of dataset.See hdl:10013/epic.39209.d002 for an overview figure. The product guide: hdl:10013/epic.39209.d012This dataset is part of the ESA Data User Element (DUE) Permafrost Full Product Set (doi:10.1594/PANGAEA.780111).

This data set consisting of initial conditions, boundary conditions and forcing profiles for the Single Column Model (SCM) version of the European Centre for Medium-range Weather Forecasts (ECMWF) model, the Integrated Forecasting System (IFS). The IFS SCM is freely available through the OpenIFS project, on application to ECMWF for a licence. The data were produced and tested for IFS CY40R1, but will be suitable for earlier model cycles, and also for future versions assuming no new boundary fields are required by a later model. The data are archived as single time-stamp maps in netCDF files. If the data are extracted at any lat-lon location and the desired timestamps concatenated (e.g. using netCDF operators), the resultant file is in the correct format for input into the IFS SCM.

The data covers the Tropical Indian Ocean/Warm Pool domain spanning 20S-20N, 42-181E. The data are available every 15 minutes from 6 April 2009 0100 UTC for a period of ten days. The total number of grid points over which an SCM can be run is 480 in the longitudinal direction, and 142 latitudinally. With over 68,000 independent grid points available for evaluation of SCM simulations, robust statistics of bias can be estimated over a wide range of boundary and climatic conditions.

The initial conditions and forcing profiles were derived by coarse-graining high resolution (4 km) simulations produced as part of the NERC Cascade project, dataset ID xfhfc (also available on CEDA). The Cascade dataset is archived once an hour. The dataset was linearly interpolated in time to produce the 15-minute resolution required by the SCM. The resolution of the coarse-grained data corresponds to the IFS T639 reduced gaussian grid (approx 32 km). The boundary conditions are as used in the operational IFS at resolution T639. The coarse graining procedure by which the data were produced is detailed in Christensen, H. M., Dawson, A. and Holloway, C. E., 'Forcing Single Column Models using High-resolution Model Simulations', in review, Journal of Advances in Modeling Earth Systems (JAMES).

For full details of the parent Cascade simulation, see Holloway et al (2012). In brief, the simulations were produced using the limited-area setup of the MetUM version 7.1 (Davies et al, 2005). The model is semi-Lagrangian and non-hydrostatic. Initial conditions were specified from the ECMWF operational analysis. A 12 km parametrised convection run was first produced over a domain 1 degree larger in each direction, with lateral boundary conditions relaxed to the ECMWF operational analysis. The 4 km run was forced using lateral boundary conditions computed from the 12 km parametrised run, via a nudged rim of 8 model grid points. The model has 70 terrain-following hybrid levels in the vertical, with vertical resolution ranging from tens of metres in the boundary layer, to 250 m in the free troposphere, and with model top at 40 km. The time step was 30 s.

The Cascade dataset did not include archived soil variables, though surface sensible and latent heat fluxes were archived. When using the dataset, it is therefore recommended that the IFS land surface scheme be deactivated and the SCM forced using the surface fluxes instead. The first day of Cascade data exhibited evidence of spin-up. It is therefore recommended that the first day be discarded, and the data used from April 7 - April 16.

The software used to produce this dataset are freely available to interested users; 1. "cg-cascade"; NCL software to produce OpenIFS forcing fields from a high-resolution MetUM simulation and necessary ECMWF boundary files. https://github.com/aopp-pred/cg-cascade Furthermore, software to facilitate the use of this dataset are also available, consisting of; 2. "scmtiles"; Python software to deploy many independent SCMs over a domain. https://github.com/aopp-pred/scmtiles 3. "openifs-scmtiles"; Python software to deploy the OpenIFS SCM using scmtiles. https://github.com/aopp-pred/openifs-scmtiles