Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
6,381
datasets available to search
ShareScore release 0.7.1
Dataset results
6,381 results for “spatial”
Spatial distribution of snow water equivalent for the Niwot Ridge, 1996 - 2019
This dataset provides a daily estimation of snow water equivalent for the Niwot Ridge during snow melting period from 1997 to 2019 at 30-meter spatial resolution. The dataset includes two series of SWE data: 1) 1996-2007 daily SWE dataset is generated by Jepsen et al., (2012); 2) 2008-2019 daily SWE dataset is generated by Dr. Kehan Yang following the same method used by Jepsen et al., (2012). In brief, a physically based reconstruction model is used to calculate daily SWE backward from snow disappearance date to peak snow accumulation. The infilled hourly climate data set for C1, Saddle and D1 (data available at https://portal.edirepository.org/nis/mapbrowse?packageid=knb-lter-nwt.168.2) is interpolated and used as the meteorological forcing in the snow energy balance calculation of SWE reconstruction. The shortwave radiation is estimated by downscaling hourly product of the Geostationary Operational Environmental Satellite (GOES) using TOPORAD tool. The USGS Landsat Level-3 fractional snow-covered area product is used to proportion potential energy flux for snowmelt at the pixel scale. Please see detailed methods included with this data package for more details and references.
The human Voice Areas: spatial organisation and inter-individual variability in temporal and extra-temporal cortices
Open the record for dataset details and reuse information.
Sediment Properties Drive Spatial Variability of Potential Methane Production and Oxidation in Small Streams
<ul> <li>This dataset contains 20 data tables (Fig.2.csv, Fig.3.csv, data_PLS_stream-main-stem.csv, Fig.4_a.csv, data_PLS_subcatch.-stream-sect.csv, Fig.4_b.csv, Fig.5.csv, Fig.S1.csv, Fig.S2.csv, Fig.S3.csv, Fig.S4.csv, Fig.S5_a.csv, Fig.S5_b.csv, Fig.S6_a.csv, Fig.S6_b.csv, Fig.S6_c.csv, Fig.S6_d.csv, TableS1_data-adjustment.csv, TableS1_lit-data.csv, PMO_surface-water.csv), we separated our data tables in the respective figures/analyses presented in our paper</li> <li>We added the units to each column title of each respective data table</li> <li>Please see "Metadata.pdf" and our paper (same title as the dataset) for more information</li> </ul> <p> </p>
Investigating dynamics between energy use and socio-demographic characteristics in spatial modeling of residential energy consumption
<p>Files represent datasets (2017 Residential Building Stock Assessment and American Community Survey 2012-2017 5-year estimate) and R-code associated with the analysis. </p>
Quantification of 3D spatial correlations between state variables and distances to the grain boundary network in full-field crystal plasticity spectral method simulations
<p>This repository provides supplementary material to our paper: <a href="https://doi.org/10.1088/1361-651X/ab7f8c">https://doi.org/10.1088/1361-651X/ab7f8c</a></p> <p><strong>DAMASKPhenoPowerLaw75x75x75TestCase.zip</strong><br> An exemplary DAMASK simulation and corresponding output, generated from DAMASK v2.0.3. We used this to debug more productively the implementation of the post-processing tools. Furthermore we employed this simulation in the paper to identify why the graph clustering grain reconstruction method in many cases fuses neighboring grains in similar orientation.</p> <p><strong>DAMASKPhenoPowerLaw256x256x256ProductionRun.zip</strong><br> All input to run the DAMASK simulation that we discussed in the paper.</p> <p><strong>DAMASKPDTSettings256x256x256ProductionRun.zip</strong><br> All damaskpdt settings files to execute the individual post-processing studies of the paper.</p> <p><strong>DAMASKPDTSlurmSubmissionScripts256x256x256ProductionRun.zip</strong><br> All SLURM scripts we used to execute the compilation of damaskpdt and post-processing on TALOS.</p> <p><strong>DAMASKPDTSlurmLogs256x256x256ProductionRun.zip</strong><br> All logs from the SLURM job management system from the individual post-processing runs.</p> <p><strong>DAMASKPDTSourceCode_USedForAnalyticalDistanceToVoronoiCellFacets.zip</strong><br> The source code to the tool we developed during the revision process of our paper to verify the methods<br> via computing analytically exact distances to the facets of the Poisson-Voronoi tessellation from the<br> DAMASK microstructure instantiation.<br> <br> <strong>DAMASKPDTSourceCode_Production.zip</strong><br> The source code we used to post-process all results from the DAMASK simulations.</p> <p><strong>GitHub repository:</strong><br> https://github.com/mkuehbach/damaskpdt</p>
Highly multiplexed histology reveals phenotypic and spatial characteristics of human Innate Lymphoid Cells in chronic inflammation - MELC tonsil data-set
<p><strong>53 marker MELC Run in human tonsil</strong>. Each image depicts the same field of view, sequentially stained with the depicted fluorescence-labelled antibodies. Images contain 2048 x 2048 pixels and are generated using an inverted wide-field fluorescence microscope with a 20x objective, a lateral resolution of 325 nm and an axial resolution above 5 µm. Images have not been normalized and intensities have not been adjusted.</p> <p> </p>
The S&M-HSTPM2d5 dataset: High Spatial-Temporal Resolution PM 2.5 Measures in Multiple Cities Sensed by Static & Mobile Devices
<p>This S&M-HSTPM2d5 dataset contains the high spatial and temporal resolution of the particulates (PM2.5) measures with the corresponding timestamp and GPS location of mobile and static devices in the three Chinese cities: Foshan, Cangzhou, and Tianjin. Different numbers of static and mobile devices were set up in each city. The sampling rate was set up as one minute in Cangzhou, and three seconds in Foshan and Tianjin. For the specific detail of the setup, please refer to the Device_Setup_Description.txt file in this repository and the data descriptor paper.</p> <p>After the data collection process, the data cleaning process was performed to remove and adjust the abnormal and drifting data. The script of the data cleaning algorithm is provided in this repository. The data cleaning algorithm only adjusts or removes individual data points. The removal of the entire device's data was done after the data cleaning algorithm with empirical judgment and graphic visualization. For specific detail of the data cleaning process, please refer to the script (Data_cleaning_algorithm.ipynb) in this repository and the data descriptor paper.</p> <p>The dataset in this repository is the processed version. The raw dataset and removed devices are not included in this repository.</p> <p>The data is stored as a CSV file. Each CSV file which is named by the device ID represents the data that was collected by the corresponding device. Each CSV file has three types of data: timestamp as the China Standard Time (GMT+8), geographic location as latitude and longitude, and PM2.5 concentration with the unit of microgram per cubic meter. The CSV files are stored in either Static or Mobile folder which represents the devices' type. The Static and Mobile folder are stored in the corresponding city's folder.</p> <p>To access the dataset, any programming language that can access CSV files is appropriate. Users can also open the CSV file directly. The get_dataset.ipynb file in this repository also provides an option of accessing the dataset. To successfully execute ipynb file, Jupyter Notebook with Python 3.0 is required. The following python library is also required:</p> <p>get_dataset.ipynb:<br> 1. os library<br> 2. pandas library</p> <p>Data_cleaning_algorithm.ipynb:<br> 1. os library<br> 2. pandas library<br> 3. datetime library<br> 4. math library</p> <p>The instruction of installing the libraries above can be found online. After installing the Jupyter Notebook with Python 3.0 and the required libraries, users can try to open the ipynb file with Jupyter Notebook and follow the instruction inside the file. </p> <p>For questions or suggestions please e-mail Xinlei Chen <xinlei.chen@sv.cmu.edu></p>
TAU-NIGENS Spatial Sound Events 2020
<p><strong>DESCRIPTION:</strong></p> <p>The <strong>TAU-NIGENS Spatial Sound Events 2020</strong> dataset contains multiple spatial sound-scene recordings, consisting of sound events of distinct categories integrated into a variety of acoustical spaces, and from multiple source directions and distances as seen from the recording position. The spatialization of all sound events is based on filtering through real spatial room impulse responses (RIRs), captured in multiple rooms of various shapes, sizes, and acoustical absorption properties. Furthermore, each scene recording is delivered in two spatial recording formats, a microphone array one (<strong>MIC</strong>), and first-order Ambisonics one (<strong>FOA</strong>). The sound events are spatialized as either stationary sound sources in the room, or moving sound sources, in which case time-variant RIRs are used. Each sound event in the sound scene is associated with a trajectory of its direction-of-arrival (DoA) to the recording point, and a temporal onset and offset time. The isolated sound event recordings used for the synthesis of the sound scenes are obtained from the <a href="https://doi.org/10.5281/zenodo.2535878">NIGENS general sound events database</a>. These recordings serve as the development dataset for the <a href="http://dcase.community/challenge2020/task-sound-event-localization-and-detection">DCASE 2020 Sound Event Localization and Detection Task</a> of the <a href="http://dcase.community/challenge2020/">DCASE 2020 Challenge</a>.</p> <p><strong>REPORT & REFERENCE:</strong></p> <p>If you use this dataset please cite the report on its creation, and the corresponding DCASE2020 task setup:</p> <p>Politis., Archontis, Adavanne, Sharath, & Virtanen, Tuomas (2020). A Dataset of Reverberant Spatial Sound Scenes with Moving Sources for Sound Event Localization and Detection. In <em>Proceedings of the Detection and Classification of Acoustic Scenes and Events 2020 Workshop (DCASE2020)</em>, Tokyo, Japan.</p> <p>A longer version with more detailed information can be also found <a href="https://arxiv.org/pdf/2006.01919.pdf">here</a>.</p> <p><strong>AIM:</strong></p> <p>The dataset includes a large number of mixtures of sound events with realistic spatial properties under different acoustic conditions, and hence it is suitable for training and evaluation of machine-listening models for sound event detection (SED), general sound source localization with diverse sounds or signal-of-interest localization, and joint sound-event-localization-and-detection (SELD). Additionally, the dataset can be used for evaluation of signal processing methods that do not necessarily rely on training, such as acoustic source localization methods and multiple-source acoustic tracking. The dataset allows evaluation of the performance and robustness of the aforementioned applications for diverse types of sounds, and under diverse acoustic conditions.</p> <p><strong>SPECIFICATIONS:</strong></p> <ul> <li>600 one-minute long sound scene recordings (development dataset).</li> <li>200 one-minute long sound scene recordings (evaluation dataset).</li> <li>Sampling rate 24kHz.</li> <li>About 700 sound event samples spread over 14 classes (see <a href="http://doi.org/10.5281/zenodo.2535878">here</a> for more details).</li> <li>8 provided cross-validation splits of 100 recordings each, with unique sound event samples and rooms in each of them.</li> <li>Two 4-channel 3-dimensional recording formats: first-order Ambisonics (<strong>FOA</strong>) and tetrahedral microphone array.</li> <li>Realistic spatialization and reverberation through RIRs collected in 15 different enclosures.</li> <li>From about 1500 to 3500 possible RIR positions across the different rooms.</li> <li>Both static reverberant and moving reverberant sound events.</li> <li>Up to two overlapping sound events allowed, temporally and spatially.</li> <li>Realistic spatial ambient noise collected from each room is added to the spatialized sound events, at varying signal-to-noise ratios (SNR) ranging from noiseless (30dB) to noisy (6dB).</li> </ul> <p>The IRs were collected in Finland by staff of Tampere University between 12/2017 - 06/2018, and between 11/2019 - 1/2020. The older measurements from five rooms were also used for the earlier <a href="https://doi.org/10.5281/zenodo.2580091">development</a> and <a href="https://doi.org/10.5281/zenodo.3066124">evaluation</a> datasets <strong>TAU Spatial Sound Events 2019</strong>, while ten additional rooms were added for this dataset. The data collection received funding from the European Research Council, grant agreement <a href="https://cordis.europa.eu/project/id/637422">637422 EVERYSOUND</a>.</p> <p>More detailed information on the dataset can be found in the included README file.</p> <p><strong>EXAMPLE APPLICATION:</strong></p> <p>An implementation of a trainable model of a convolutional recurrent neural network, performing joint SELD, trained and evaluated with this dataset is provided <a href="https://github.com/sharathadavanne/seld-dcase2020">here</a>. This implementation serves as the baseline method in the <a href="http://dcase.community/challenge2020/task-sound-event-localization-and-detection">DCASE 2020 Sound Event Localization and Detection Task</a>.</p> <p><strong>DEVELOPMENT AND EVALUATION:</strong></p> <p>Version 1.0 of the dataset included only the 600 development audio recordings and labels, used by the participants of Task 3 of DCASE2020 Challenge to train and validate their submitted systems. Version 1.1 included additionally the 200 evaluation audio recordings without labels, for the evaluation phase of DCASE2020. The latest version 1.2, published after the completion of the challenge, includes also the labels for the evaluation files.</p> <p>If researchers wish to compare their system against the submissions of DCASE2020 Challenge, they will have directly comparable results if they use the evaluation data as their testing set.</p> <p><strong>DOWNLOAD INSTRUCTIONS:</strong></p> <p>The three files, <strong><em>foa_dev.z01</em></strong>,<strong><em> foa_dev.z02</em></strong>, and <strong><em>foa_dev.zip</em></strong>, correspond to audio data of the <strong>FOA </strong>recording format.<br> The three files, <strong><em>mic_dev.z01</em></strong>,<strong><em> mic_dev.z02</em></strong>, and <strong><em>mic_dev.zip</em></strong>, correspond to audio data of the <strong>MIC</strong> recording format.<br> The <strong><em>metadata_dev.zip</em></strong> is the common metadata for both formats.</p> <p>The file, <em><strong>foa_eval.zip</strong></em>, corresponds to audio data of the <strong>FOA</strong> recording format for the evaluation dataset.<br> The file, <em><strong>mic_eval.zip</strong></em>, corresponds to audio data of the <strong>MIC</strong> recording format for the evaluation dataset.<br> The <em><strong>metadata_eval.zip</strong></em> is the common metadata for both formats. An info file is included (<em>metadata_eval_info.txt</em>) which specifies which of the two evaluation folds the mix file belongs to, and what is its number of overlapping events.</p> <p>Download the zip files corresponding to the format of interest and use your favorite compression tool to unzip these split zip files. To extract a split zip archive (named as zip, z01, z02, ...), you could use, for example, the following syntax in Linux or OSX terminal:</p> <ol> <li>Combine the split archive to a single archive: <pre>zip -s 0 split.zip --out single.zip</pre> </li> <li>Extract the single archive using unzip: <pre>unzip single.zip</pre> </li> </ol>
Soil organic carbon stocks and trends (1984-2019) predicted at 30m spatial resolution for topsoil in natural areas of South Africa
<p>Link to scientific publication: <a href="https://doi.org/10.1016/j.scitotenv.2021.145384">https://doi.org/10.1016/j.scitotenv.2021.145384</a></p> <p>Soil organic carbon (SOC) stocks (kg C m-2) are predicted over natural areas (excluding water, urban, and cultivated) of South Africa using a machine learning workflow driven by optical satellite data and other ancillary climatic, morphometric and biological covariates. The temporal scope covers 1984-2019. The spatial scope covers 0-30cm topsoil in South Africa natural land area (84% of the country). See methodology in linked publication for details. Data are provided here at 30m spatial resolution in GeoTIFF files. There is a dataset for the long-term average SOC and trend in SOC. Each dataset is split into four files (suffix *_1, *_2 etc.) covering separate regions of South Africa for ease of download. The raster files are:</p> <ul> <li>"SOC_mean_30m..." - average of annual SOC predictions between 1984 and 2019. Values are expressed in kg C m-2</li> <li>"SOC_trend_30m..." - long-term trend in SOC derived from the Sens slope (M) across annual SOC values between 1984 and 2019. Pixel values (Y) are expressed as a percentage change over the 35 years relative to the long-term mean (X). Y = M / X * 100 * 35 years</li> </ul> <p>NB: All files are scaled by *100 and converted to floating data point to save space. To back-convert to original values, simply divide the raster values by 100.</p>
Spatially Resolved Infrared Radiofluorescence (SR IR-RF) Image Data
<p>This dataset contains measurement sequences and data output <br> of spatially resolved infrared radiofluorescence (SR IR-RF) measurements<br> on K-feldspar samples carried out at the IRAMAT-CRP2A, UMR 5060, CNRS-Université Bordeaux Montaigne (France)<br> in 2019. The data analysis was performed in 2020. </p> <p>The data may serve as reference data and allow detailed inspection by others to <br> verify or advance the used analysis procedures. </p> <p>Along with the raw image data (TIF-files), the datasets also contain documented R scripts used for data processing and partly treated data as an example. To reproduce the full data analysis, additional software is needed; not part of this repository. </p> <p>Further details can be found in the README.md (README.html), which is part of the dataset.</p>
Spatial distribution data set of wetlands in Baiyangdian Basin
<p>As one of the wetland systems in the northern plain of China, Baiyangdian plays a key role in ensuring the water resources security and good ecological environment of Xiong'an New Area. Understanding the current situation of the wetland ecosystem in Baiyangdian basin is also of great significance for the construction of the New Area and future scientific planning. Based on the 10 meter spatial resolution sentinel-2B image provided by ESA in September 2017, combined with Google Earth high resolution satellite image (resolution 0.23m), the network distribution map and water system distribution map of Baiyangdian basin wetland ecosystem in 2017 were drawn by artificial visual interpretation and machine automatic classification It provides the basis for the study of the connectivity (including hydrological connectivity and landscape connectivity).</p> <p>The boundary of Baiyangdian basin in this data set is from the basic geographic information map of Baiyangdian basin provided by Zhou Wei and others. The DEM is the GDEM digital elevation data with 30m resolution. The original image data of wetland remote sensing classification comes from the sentinel-2b remote sensing image provided by ESA on September 20, 2017. This data set uses the second, third, fourth and eighth bands of 10 meter resolution in the image, carries out radiation calibration, mosaic, mosaic and other preprocessing operations in SNAP and ArcGIS 10.2 software, and carries out supervised classification in ENVI 5.3 software. The data used for river channel extraction is based on Google Earth high resolution satellite images.</p> <p>The research and development steps of this dataset include: preprocessing sentinel-2B image, establishing wetland classification system and selecting samples, mapping the latest wetland ecosystem network distribution map of Baiyangdian basin by support vector machine classification; obtaining river network of Baiyangdian basin by visual interpretation based on Google Earth high resolution satellite image (resolution 0.23m).</p> <p>The spatial distribution data set of Baiyangdian Wetland includes vector data and raster data: (1) Baiyangdian basin boundary data (. SHP); Baiyangdian basin river network data (. shp); (2) Baiyangdian basin land use / cover classification data (including the classification data of the study area and the river 3 km buffer) (. tif); Baiyangdian basin constructed wetland and natural wetland distribution map (. shp); Baiyangdian basin slope map (. tif).</p> <p>According to the river network map of Baiyangdian basin obtained by manual visual interpretation, the total length of the river in Baiyangdian basin is about 2440 km and the total area is 514 km2. Among them, there are 177 km2 river channels in mountainous area, 866 km in length, distributed in Northeast southwest direction, mostly at the junction of forest land and cultivated land; and 337 km2 river channels in plain area, 1574 km in length.</p> <p>Baiyangdian basin is divided into eight types of land use / cover: river, flood plain, lake, marsh, ditch, cultivated land, forest land and construction land. The remote sensing monitoring results show that the wetland area of Baiyangdian basin accounted for 13.90 % in 2017. Among all wetland types, the area of marsh is the largest, followed by the area of flood plain, ditch accounts for about 1%, and the proportion of lake and river is less than 0.5%. Combined with the land use / cover classification map and the distribution of slope and elevation, it can be seen that nearly 60% of the area of woodland is distributed in 10 ° to 30 ° mountain area, and the rest of the land use / cover types are mainly distributed in 0 ° to 2 ° area. The elevation statistics show that nearly 80% of the lakes and large reservoirs are distributed in the height of 100 m to 300 m, the distribution of marsh is relatively uniform, mainly in the high altitude area of 20 m to 300 m, the types of construction land, flood area and cultivated land are mainly concentrated in the area of 20 m to 100 m, and rivers and ditches are mainly concentrated in the area of 0 m to 100 m.</p> <p>Based on the classification results of land use / cover within the river, it can be found that the main land use type is wetland. Specifically, the types of swamp, flood area and lake are the most, while the types of ditch and river are less. With the increase of the buffer area, the proportion of non wetland type gradually increased, while the proportion of wetland type gradually decreased. The main wetland types in 1-3km buffer zone on both sides of the river are swamp and flood zone. It is worth noting that nearly one third of the River belongs to cultivated land, that is, the river occupation is serious. In terms of area, about 1 / 3 rivers and 3 / 4 lakes are distributed in the river course. Most of the water bodies in the river course are controlled by human beings, but the marsh area in the river course only accounts for about 3% of the marsh area in the whole river course.</p> <p>River occupation will not only directly reduce the connectivity of wetlands in the basin, but also cause some environmental and economic problems such as water pollution. However, if the connectivity of wetlands is reduced, the ecological and environmental functions of wetlands will be destroyed, which will pose a great threat to the water security of the basin. Taking Baiyangdian basin as a whole, improving the connectivity of wetlands and enhancing the ecological and environmental functions of wetlands in the basin will help to improve the water ecological and environmental security of xiong'an new area and Baiyangdian basin.</p>
Sudden_Oak_Death_in_Oregon_Forests: Spatial and temporal population dynamics of the sudden oak death epidemic in Oregon Forests
<p>Release of code associated with the submitted manuscript</p> <p><strong>Authors</strong></p> <p>ZN Kamvar, MM Larsen, AM Kanaskie, EM Hansen, and NJ Grünwald.</p> <p><strong>Title</strong></p> <p>Spatial and temporal population dynamics of the sudden oak death epidemic in Oregon Forests.</p>
Object-based audio scene files for variations of the spatial arrangement in pop mixes for Wave Field Synthesis
<p>This entry contains object-based audio meta-data to generate the mixes published at http://dx.doi.org/10.5281/zenodo.61000.</p> <p>Have a look at README.md for further details.</p>
Spatial distributions of Solar Energetic Particle Fe ions at t=12 h
<p>These images are supplementary to Figure 2 of the publication entitled "Solar Energetic Particle drifts and the energy dependence of 1 AU charge states" by S. Dalla, M.S. Marsh, M. Battarbee, accepted by Astrophysical Journal (2016). The paper is also available at https://arxiv.org/abs/1610.05104.</p> <p>In each figure contour plots of locations of Fe ions of charges Q=20, 16, 12 and 8 at time t=12 hr are shown. The first image gives z vs r_xy projections and the second figure y vs x projections, with only particles within 20 degrees of the heliographic equator included in the x-y projections.</p>
Spatially-localized X-ray scattering and X-ray microtomography measurements on Moso bamboo
<p><strong>Spatially-localized X-ray scattering and X-ray microtomography measurements on Moso bamboo</strong></p> <p> </p> <p>This data set is originally used in:</p> <p>Ahvenainen, P., Dixon, P. G., Kallonen, A., Suhonen, H., Gibson, L. J., & Svedström, K. (2017). Spatially-localized bench-top X-ray scattering reveals tissue-specific microfibril orientation in Moso bamboo. <em>Plant Methods</em>. <strong>13</strong>:5 DOI: 10.1186/s13007-016-0155-1</p> <p>This data set includes measurements on Moso bamboo (<em>Phyllostachys edulis</em>) performed with two separate set-ups at the Department of Physics, University of Helsinki as described in the above open-access publication. The X-ray microtomography (XMT) measurements, X-ray diffraction tomography (XDT) and localized X-ray scattering (LXS) are done with set-up 1. In LXS, the region-of-interest is selected from a tomographic reconstruction slice based on the XMT measurement using a small X-ray beam (diameter: 200 µm). Additional wide-angle X-ray scattering (WAXS) measurements are conducted with set-up 2 using a larger X-ray beam (diameter approx. 1 mm). </p> <p>The two-dimensional scattering patterns (Pilatus 1M hybrid pixel array detector) and tomographic reconstruction slices obtained with set-up 2 are stored as TIFF-images (.tif). The two-dimensional scattering patterns (MAR345 image plate detector) obtained with set-up 2 are stored as 32-bit RAW files (unsigned integers, 2300 columns, 2300 rows). </p> <p>The novel combined WAXS/XMT set up (set-up 1) is first presented in: Suuronen, J.-P., Kallonen, A., Hänninen, V., Blomberg, M., Hämäläinen, K., & Serimaa, R. (2014). Bench-top X-ray microtomography complemented with spatially localized X-ray scattering experiments. <em>Journal of Applied Crystallography</em>, <strong>47</strong>(1), 471–475. doi:10.1107/S1600576713031105</p> <p>Any queries related to the data set or the related Plant Methods article may be directed to the first author by email:</p> <p>Patrik Ahvenainen, PhD; patrik.ahvenainen@alumni.helsinki.fi</p>
Spatial impulse wave
<p>ENGLISH<br> <br> Landslides and avalanches in natural lakes and reservoirs may generate so-called impulse waves. The run-up effects of these waves at the shore are similar to those of tsunamis. Hydraulic experiments in the laboratory help to estimate key wave characteristics, including the wave height and celerity. The picture presents two images of an experiment in the wave basin of the Laboratory of Hydraulics, Hydrology and Glaciology (VAW) at ETH Zurich. At the upper left, the moment shortly after the slide has hit the water surface is shown. The slide transfers its kinetic energy to the water and generates a wave. In the section on the lower right, the waves have propagated circularly from the impact location. The water is dyed white so that a grid can be projected onto the surface. During the experiment, this raster projection is simultaneously filmed by several cameras and the analysis of the image data allows for an accurate determination of the wave height decay.<br> <br> GERMAN<br> <br> Erdrutsche und Lawinen in natürliche Seen oder Stauseen können sogenannte Impulswellen auslösen. Die Auswirkungen beim Auflaufen dieser Wellen am Ufer sind mit denen von Tsunamis vergleichbar. Hydraulische Experimente im Labor helfen dabei, massgebliche Welleneigenschaften, wie beispielsweise die Höhe oder die Ausbreitungsgeschwindigkeit, abzuschätzen. Das Bild stellt zwei Aufnahmen eines Experiments im Wellenbecken der Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie der ETH Zürich zu unterschiedlichen Zeitpunkten dar. Oben links ist der Moment kurz nach dem Auftreffen des Rutsches auf die Wasseroberfläche zu sehen. Der Rutsch überträgt dabei seine Bewegungsenergie auf das Wasser und erzeugt eine Welle. Im Ausschnitt unten rechts haben sich die Wellen kreisförmig von der Eintauchstelle weg ausgebreitet. Das Wasser ist weiss eingefärbt, damit ein Raster auf die Oberfläche projiziert werden kann. Diese Rasterprojektion wird während des Experiments gleichzeitig von mehreren Kameras gefilmt und die Auswertung der Bilddaten ermöglicht eine genaue Bestimmung der Wellenhöhenabnahme.</p>
Global Pasture Watch - Annual grassland class and extent maps at 30-m spatial resolution (2000—2024)
<h2><strong>Sub-dataset: Dominant grassland class, 2018-2020</strong></h2> <h2>Description</h2> <p>Global annual grassland class and extent for 2000—2024 produced by <a href="https://doi.org/10.1038/s41597-024-04139-6">Parente et al. (2024)</a> within the scope of the <a href="https://landcarbonlab.org/data/global-grassland-and-livestock-monitoring/">Global Pasture Watch initiative</a>. The mapped grassland extent includes any land cover type, which contains at least <strong>30% of dry or wet low vegetation</strong>, dominated by grasses and forbs (less than 3 meters) and a:</p> <ul> <li>maximum of 50% tree canopy cover (greater than 5 meters),</li> <li>maximum of 70% of other woody vegetation (scrubs and open shrubland), and</li> <li>maximum of 50% active cropland cover in mosaic landscapes of cropland & other vegetation.</li> </ul> <p>The grassland extent is classified into two classes:</p> <ul> <li><strong>Cultivated grassland</strong>: Areas where grasses and other forage plants have been intentionally planted and managed, as well as areas of native grassland-type vegetation where they clearly exhibit active and 'heavy' management for specific human-directed uses, such as directed grazing of livestock.</li> <li><strong>Natural/semi-natural grassland</strong>: Relatively undisturbed native grasslands/short-height vegetation, such as steppes and tundra, as well as areas that have experienced varying degrees of human activity in the past, which may contain a mix of native and introduced species due to historical land use and natural processes. In general, they exhibit natural-looking patterns of varied vegetation and clearly ordered hydrological relationships throughout the landscape.</li> <li><strong>Open shrubland (v2-beta): </strong>Land on which the vegetation is dominated by low-growing woody plants, characterized by a sparse distribution of shrubs and dominated by woody perennials. Typically covers 50—75% of the area, with significant open ground (with or without herbaceous understory) between them, where shrub canopies are less than 10 meters in diameter, and tree cover is below 10%, meaning they do not form a continuous or semi-continuous canopy.</li> </ul> <p>The dataset is organized in 69 global mosaics (25 years for each time series) in COG (Cloud Optimized GeoTIFF) format, WGS84 Coordinate Systems (EPSG:4326) and pixel size equal to 0.00025 degrees, including:</p> <ul> <li><strong>Probabilities</strong> of cultivated grassland (values range from 0–100),</li> <li><strong>Probabilities</strong> of natural/semi-natural grassland (values range from 0–100), and</li> <li><strong>Probabilities</strong> of open shrubland (values range from 0–100), and</li> <li><strong>Dominant</strong> class (0-other land cover, 1-cultivated grassland and 2-natural/semi-natural grassland, 3-open shrubland).</li> </ul> <p>All raster files are in unsigned <code>8-bit integer format</code> and use <code>255</code> as no-data value (pixels ignored by prediction), following an specific naming convention:</p> <ol> <li>Project name: Global Pasture Watch (<code>gpw</code>)</li> <li>Class name: cultivated grassland (<code>cultiv.grassland</code>), natural/semi-natural grassland (<code>nat.semi.grassland</code>), open shrubland (<code>open.shrubland</code>) and dominant grassland (<code>grassland</code>)</li> <li>Procedure combination: Random Forest (<code>rf</code>), median filter (med.filt) and balanced threshold (<code>bthr</code>).</li> <li>Variable type: probability (<code>p</code>) and factor class (<code>c</code>)</li> <li>Spatial resolution: 30m</li> <li>Begin of time reference: date of first Landsat composite used by the modeling (<code>20240101</code>)</li> <li>End of time reference: date of last Landsat composite used by the modeling (<code>20241231</code>)</li> <li>Spatial extent: global (<code>go</code>)</li> <li>Coordinate system: World Geodetic System 1984, used in GPS (<code>epsg.4326</code>)</li> <li>Version: v2</li> </ol> <h3>Related resources</h3> <ul> <li><strong>Maps of dominant grassland:</strong><br><a href="http://doi.org/10.5281/zenodo.15646181">2000-2002</a><a href="http://doi.org/10.5281/zenodo.15644486"> 2003-2005</a><a href="http://doi.org/10.5281/zenodo.15644623"> 2006-2008</a><a href="http://doi.org/10.5281/zenodo.15647042"> 2009-2011</a><a href="http://doi.org/10.5281/zenodo.15647681"> 2012-2014</a><a href="http://doi.org/10.5281/zenodo.15648306"> 2015-2017</a><a href="http://doi.org/10.5281/zenodo.15648551"> 2018-2020</a><a href="http://doi.org/10.5281/zenodo.15648751"> 2021-2023</a> <a href="http://doi.org/10.5281/zenodo.15649332">2024</a></li> <li><strong>Probability maps of cultivated grassland:</strong><br><a href="https://zenodo.org/records/13890401/files/ggc-30m.csv?download=1">2000-2024 (All URLs)</a></li> <li><strong>Probability maps of natural/semi-natural grassland:</strong><br><a href="https://zenodo.org/records/13890401/files/ggc-30m.csv?download=1">2000-2024 (All URLs)</a></li> <li> <div><strong>Grassland reference samples based on VHR imagery (2000–2024):</strong><br><a href="https://doi.org/10.5281/zenodo.15631655">GeoPackage files</a></div> </li> <li><strong>Global machine learning models (Random Forest):</strong><br><a href="https://doi.org/10.5281/zenodo.13952806">Parquet and joblib python files</a></li> <li><strong>Reference sampling design derived by FSCV:</strong><br><a href="https://doi.org/10.5281/zenodo.11391517">GeoPackage and raster files</a></li> <li><strong>Harmonized reference samples based on existing LULC dataset:</strong><br><a href="https://doi.org/10.5281/zenodo.13951976">GeoPackage and raster files</a></li> <li><strong>Source code for reproducibility:<br></strong><a href="https://doi.org/10.5281/zenodo.13952867">GitHub release</a><strong><br></strong></li> <li><strong>Mapping feedback tool:</strong><br><a href="https://geo-wiki.org">GeoWiki</a></li> <li><strong>Data catalogues:</strong><br><a href="https://stac.openlandmap.org/gpw_ggc-30m/collection.json?.language=en">OpenLandMap STAC</a> <a href="https://global-pasture-watch.projects.earthengine.app/view/ggc-30m">Google Earth Engine</a></li> </ul> <p><strong>Support</strong></p> <p>For questions of bugs/inconsistencies related to the dataset raise a GitHub issue in <a href="https://github.com/wri/global-pasture-watch">https://github.com/wri/global-pasture-watch</a></p>
Global Pasture Watch - Annual grassland class and extent maps at 30-m spatial resolution (2000—2024)
<h2><strong>Sub-dataset: Dominant grassland class, 2021-2023</strong></h2> <h2>Description</h2> <p>Global annual grassland class and extent for 2000—2024 produced by <a href="https://doi.org/10.1038/s41597-024-04139-6">Parente et al. (2024)</a> within the scope of the <a href="https://landcarbonlab.org/data/global-grassland-and-livestock-monitoring/">Global Pasture Watch initiative</a>. The mapped grassland extent includes any land cover type, which contains at least <strong>30% of dry or wet low vegetation</strong>, dominated by grasses and forbs (less than 3 meters) and a:</p> <ul> <li>maximum of 50% tree canopy cover (greater than 5 meters),</li> <li>maximum of 70% of other woody vegetation (scrubs and open shrubland), and</li> <li>maximum of 50% active cropland cover in mosaic landscapes of cropland & other vegetation.</li> </ul> <p>The grassland extent is classified into two classes:</p> <ul> <li><strong>Cultivated grassland</strong>: Areas where grasses and other forage plants have been intentionally planted and managed, as well as areas of native grassland-type vegetation where they clearly exhibit active and 'heavy' management for specific human-directed uses, such as directed grazing of livestock.</li> <li><strong>Natural/semi-natural grassland</strong>: Relatively undisturbed native grasslands/short-height vegetation, such as steppes and tundra, as well as areas that have experienced varying degrees of human activity in the past, which may contain a mix of native and introduced species due to historical land use and natural processes. In general, they exhibit natural-looking patterns of varied vegetation and clearly ordered hydrological relationships throughout the landscape.</li> <li><strong>Open shrubland (v2-beta): </strong>Land on which the vegetation is dominated by low-growing woody plants, characterized by a sparse distribution of shrubs and dominated by woody perennials. Typically covers 50—75% of the area, with significant open ground (with or without herbaceous understory) between them, where shrub canopies are less than 10 meters in diameter, and tree cover is below 10%, meaning they do not form a continuous or semi-continuous canopy.</li> </ul> <p>The dataset is organized in 69 global mosaics (25 years for each time series) in COG (Cloud Optimized GeoTIFF) format, WGS84 Coordinate Systems (EPSG:4326) and pixel size equal to 0.00025 degrees, including:</p> <ul> <li><strong>Probabilities</strong> of cultivated grassland (values range from 0–100),</li> <li><strong>Probabilities</strong> of natural/semi-natural grassland (values range from 0–100), and</li> <li><strong>Probabilities</strong> of open shrubland (values range from 0–100), and</li> <li><strong>Dominant</strong> class (0-other land cover, 1-cultivated grassland and 2-natural/semi-natural grassland, 3-open shrubland).</li> </ul> <p>All raster files are in unsigned <code>8-bit integer format</code> and use <code>255</code> as no-data value (pixels ignored by prediction), following an specific naming convention:</p> <ol> <li>Project name: Global Pasture Watch (<code>gpw</code>)</li> <li>Class name: cultivated grassland (<code>cultiv.grassland</code>), natural/semi-natural grassland (<code>nat.semi.grassland</code>), open shrubland (<code>open.shrubland</code>) and dominant grassland (<code>grassland</code>)</li> <li>Procedure combination: Random Forest (<code>rf</code>), median filter (med.filt) and balanced threshold (<code>bthr</code>).</li> <li>Variable type: probability (<code>p</code>) and factor class (<code>c</code>)</li> <li>Spatial resolution: 30m</li> <li>Begin of time reference: date of first Landsat composite used by the modeling (<code>20240101</code>)</li> <li>End of time reference: date of last Landsat composite used by the modeling (<code>20241231</code>)</li> <li>Spatial extent: global (<code>go</code>)</li> <li>Coordinate system: World Geodetic System 1984, used in GPS (<code>epsg.4326</code>)</li> <li>Version: v2</li> </ol> <h3>Related resources</h3> <ul> <li><strong>Maps of dominant grassland:</strong><br><a href="http://doi.org/10.5281/zenodo.15646181">2000-2002</a><a href="http://doi.org/10.5281/zenodo.15644486"> 2003-2005</a><a href="http://doi.org/10.5281/zenodo.15644623"> 2006-2008</a><a href="http://doi.org/10.5281/zenodo.15647042"> 2009-2011</a><a href="http://doi.org/10.5281/zenodo.15647681"> 2012-2014</a><a href="http://doi.org/10.5281/zenodo.15648306"> 2015-2017</a><a href="http://doi.org/10.5281/zenodo.15648551"> 2018-2020</a><a href="http://doi.org/10.5281/zenodo.15648751"> 2021-2023</a> <a href="http://doi.org/10.5281/zenodo.15649332">2024</a></li> <li><strong>Probability maps of cultivated grassland:</strong><br><a href="https://zenodo.org/records/13890401/files/ggc-30m.csv?download=1">2000-2024 (All URLs)</a></li> <li><strong>Probability maps of natural/semi-natural grassland:</strong><br><a href="https://zenodo.org/records/13890401/files/ggc-30m.csv?download=1">2000-2024 (All URLs)</a></li> <li> <div><strong>Grassland reference samples based on VHR imagery (2000–2024):</strong><br><a href="https://doi.org/10.5281/zenodo.15631655">GeoPackage files</a></div> </li> <li><strong>Global machine learning models (Random Forest):</strong><br><a href="https://doi.org/10.5281/zenodo.13952806">Parquet and joblib python files</a></li> <li><strong>Reference sampling design derived by FSCV:</strong><br><a href="https://doi.org/10.5281/zenodo.11391517">GeoPackage and raster files</a></li> <li><strong>Harmonized reference samples based on existing LULC dataset:</strong><br><a href="https://doi.org/10.5281/zenodo.13951976">GeoPackage and raster files</a></li> <li><strong>Source code for reproducibility:<br></strong><a href="https://doi.org/10.5281/zenodo.13952867">GitHub release</a><strong><br></strong></li> <li><strong>Mapping feedback tool:</strong><br><a href="https://geo-wiki.org">GeoWiki</a></li> <li><strong>Data catalogues:</strong><br><a href="https://stac.openlandmap.org/gpw_ggc-30m/collection.json?.language=en">OpenLandMap STAC</a> <a href="https://global-pasture-watch.projects.earthengine.app/view/ggc-30m">Google Earth Engine</a></li> </ul> <p><strong>Support</strong></p> <p>For questions of bugs/inconsistencies related to the dataset raise a GitHub issue in <a href="https://github.com/wri/global-pasture-watch">https://github.com/wri/global-pasture-watch</a></p>
Global Pasture Watch - Annual grassland class and extent maps at 30-m spatial resolution (2000—2024)
<h2><strong>Sub-dataset: Dominant grassland class, 2015-2017</strong></h2> <h2>Description</h2> <p>Global annual grassland class and extent for 2000—2024 produced by <a href="https://doi.org/10.1038/s41597-024-04139-6">Parente et al. (2024)</a> within the scope of the <a href="https://landcarbonlab.org/data/global-grassland-and-livestock-monitoring/">Global Pasture Watch initiative</a>. The mapped grassland extent includes any land cover type, which contains at least <strong>30% of dry or wet low vegetation</strong>, dominated by grasses and forbs (less than 3 meters) and a:</p> <ul> <li>maximum of 50% tree canopy cover (greater than 5 meters),</li> <li>maximum of 70% of other woody vegetation (scrubs and open shrubland), and</li> <li>maximum of 50% active cropland cover in mosaic landscapes of cropland & other vegetation.</li> </ul> <p>The grassland extent is classified into two classes:</p> <ul> <li><strong>Cultivated grassland</strong>: Areas where grasses and other forage plants have been intentionally planted and managed, as well as areas of native grassland-type vegetation where they clearly exhibit active and 'heavy' management for specific human-directed uses, such as directed grazing of livestock.</li> <li><strong>Natural/semi-natural grassland</strong>: Relatively undisturbed native grasslands/short-height vegetation, such as steppes and tundra, as well as areas that have experienced varying degrees of human activity in the past, which may contain a mix of native and introduced species due to historical land use and natural processes. In general, they exhibit natural-looking patterns of varied vegetation and clearly ordered hydrological relationships throughout the landscape.</li> <li><strong>Open shrubland (v2-beta): </strong>Land on which the vegetation is dominated by low-growing woody plants, characterized by a sparse distribution of shrubs and dominated by woody perennials. Typically covers 50—75% of the area, with significant open ground (with or without herbaceous understory) between them, where shrub canopies are less than 10 meters in diameter, and tree cover is below 10%, meaning they do not form a continuous or semi-continuous canopy.</li> </ul> <p>The dataset is organized in 69 global mosaics (25 years for each time series) in COG (Cloud Optimized GeoTIFF) format, WGS84 Coordinate Systems (EPSG:4326) and pixel size equal to 0.00025 degrees, including:</p> <ul> <li><strong>Probabilities</strong> of cultivated grassland (values range from 0–100),</li> <li><strong>Probabilities</strong> of natural/semi-natural grassland (values range from 0–100), and</li> <li><strong>Probabilities</strong> of open shrubland (values range from 0–100), and</li> <li><strong>Dominant</strong> class (0-other land cover, 1-cultivated grassland and 2-natural/semi-natural grassland, 3-open shrubland).</li> </ul> <p>All raster files are in unsigned <code>8-bit integer format</code> and use <code>255</code> as no-data value (pixels ignored by prediction), following an specific naming convention:</p> <ol> <li>Project name: Global Pasture Watch (<code>gpw</code>)</li> <li>Class name: cultivated grassland (<code>cultiv.grassland</code>), natural/semi-natural grassland (<code>nat.semi.grassland</code>), open shrubland (<code>open.shrubland</code>) and dominant grassland (<code>grassland</code>)</li> <li>Procedure combination: Random Forest (<code>rf</code>), median filter (med.filt) and balanced threshold (<code>bthr</code>).</li> <li>Variable type: probability (<code>p</code>) and factor class (<code>c</code>)</li> <li>Spatial resolution: 30m</li> <li>Begin of time reference: date of first Landsat composite used by the modeling (<code>20240101</code>)</li> <li>End of time reference: date of last Landsat composite used by the modeling (<code>20241231</code>)</li> <li>Spatial extent: global (<code>go</code>)</li> <li>Coordinate system: World Geodetic System 1984, used in GPS (<code>epsg.4326</code>)</li> <li>Version: v2</li> </ol> <h3>Related resources</h3> <ul> <li><strong>Maps of dominant grassland:</strong><br><a href="http://doi.org/10.5281/zenodo.15646181">2000-2002</a><a href="http://doi.org/10.5281/zenodo.15644486"> 2003-2005</a><a href="http://doi.org/10.5281/zenodo.15644623"> 2006-2008</a><a href="http://doi.org/10.5281/zenodo.15647042"> 2009-2011</a><a href="http://doi.org/10.5281/zenodo.15647681"> 2012-2014</a><a href="http://doi.org/10.5281/zenodo.15648306"> 2015-2017</a><a href="http://doi.org/10.5281/zenodo.15648551"> 2018-2020</a><a href="http://doi.org/10.5281/zenodo.15648751"> 2021-2023</a> <a href="http://doi.org/10.5281/zenodo.15649332">2024</a></li> <li><strong>Probability maps of cultivated grassland:</strong><br><a href="https://zenodo.org/records/13890401/files/ggc-30m.csv?download=1">2000-2024 (All URLs)</a></li> <li><strong>Probability maps of natural/semi-natural grassland:</strong><br><a href="https://zenodo.org/records/13890401/files/ggc-30m.csv?download=1">2000-2024 (All URLs)</a></li> <li> <div><strong>Grassland reference samples based on VHR imagery (2000–2024):</strong><br><a href="https://doi.org/10.5281/zenodo.15631655">GeoPackage files</a></div> </li> <li><strong>Global machine learning models (Random Forest):</strong><br><a href="https://doi.org/10.5281/zenodo.13952806">Parquet and joblib python files</a></li> <li><strong>Reference sampling design derived by FSCV:</strong><br><a href="https://doi.org/10.5281/zenodo.11391517">GeoPackage and raster files</a></li> <li><strong>Harmonized reference samples based on existing LULC dataset:</strong><br><a href="https://doi.org/10.5281/zenodo.13951976">GeoPackage and raster files</a></li> <li><strong>Source code for reproducibility:<br></strong><a href="https://doi.org/10.5281/zenodo.13952867">GitHub release</a><strong><br></strong></li> <li><strong>Mapping feedback tool:</strong><br><a href="https://geo-wiki.org">GeoWiki</a></li> <li><strong>Data catalogues:</strong><br><a href="https://stac.openlandmap.org/gpw_ggc-30m/collection.json?.language=en">OpenLandMap STAC</a> <a href="https://global-pasture-watch.projects.earthengine.app/view/ggc-30m">Google Earth Engine</a></li> </ul> <p><strong>Support</strong></p> <p>For questions of bugs/inconsistencies related to the dataset raise a GitHub issue in <a href="https://github.com/wri/global-pasture-watch">https://github.com/wri/global-pasture-watch</a></p>
Global Pasture Watch - Annual grassland class and extent maps at 30-m spatial resolution (2000—2024)
<h2><strong>Sub-dataset: Dominant grassland class, 2009-2011</strong></h2> <h2>Description</h2> <p>Global annual grassland class and extent for 2000—2024 produced by <a href="https://doi.org/10.1038/s41597-024-04139-6">Parente et al. (2024)</a> within the scope of the <a href="https://landcarbonlab.org/data/global-grassland-and-livestock-monitoring/">Global Pasture Watch initiative</a>. The mapped grassland extent includes any land cover type, which contains at least <strong>30% of dry or wet low vegetation</strong>, dominated by grasses and forbs (less than 3 meters) and a:</p> <ul> <li>maximum of 50% tree canopy cover (greater than 5 meters),</li> <li>maximum of 70% of other woody vegetation (scrubs and open shrubland), and</li> <li>maximum of 50% active cropland cover in mosaic landscapes of cropland & other vegetation.</li> </ul> <p>The grassland extent is classified into two classes:</p> <ul> <li><strong>Cultivated grassland</strong>: Areas where grasses and other forage plants have been intentionally planted and managed, as well as areas of native grassland-type vegetation where they clearly exhibit active and 'heavy' management for specific human-directed uses, such as directed grazing of livestock.</li> <li><strong>Natural/semi-natural grassland</strong>: Relatively undisturbed native grasslands/short-height vegetation, such as steppes and tundra, as well as areas that have experienced varying degrees of human activity in the past, which may contain a mix of native and introduced species due to historical land use and natural processes. In general, they exhibit natural-looking patterns of varied vegetation and clearly ordered hydrological relationships throughout the landscape.</li> <li><strong>Open shrubland (v2-beta): </strong>Land on which the vegetation is dominated by low-growing woody plants, characterized by a sparse distribution of shrubs and dominated by woody perennials. Typically covers 50—75% of the area, with significant open ground (with or without herbaceous understory) between them, where shrub canopies are less than 10 meters in diameter, and tree cover is below 10%, meaning they do not form a continuous or semi-continuous canopy.</li> </ul> <p>The dataset is organized in 69 global mosaics (25 years for each time series) in COG (Cloud Optimized GeoTIFF) format, WGS84 Coordinate Systems (EPSG:4326) and pixel size equal to 0.00025 degrees, including:</p> <ul> <li><strong>Probabilities</strong> of cultivated grassland (values range from 0–100),</li> <li><strong>Probabilities</strong> of natural/semi-natural grassland (values range from 0–100), and</li> <li><strong>Probabilities</strong> of open shrubland (values range from 0–100), and</li> <li><strong>Dominant</strong> class (0-other land cover, 1-cultivated grassland and 2-natural/semi-natural grassland, 3-open shrubland).</li> </ul> <p>All raster files are in unsigned <code>8-bit integer format</code> and use <code>255</code> as no-data value (pixels ignored by prediction), following an specific naming convention:</p> <ol> <li>Project name: Global Pasture Watch (<code>gpw</code>)</li> <li>Class name: cultivated grassland (<code>cultiv.grassland</code>), natural/semi-natural grassland (<code>nat.semi.grassland</code>), open shrubland (<code>open.shrubland</code>) and dominant grassland (<code>grassland</code>)</li> <li>Procedure combination: Random Forest (<code>rf</code>), median filter (med.filt) and balanced threshold (<code>bthr</code>).</li> <li>Variable type: probability (<code>p</code>) and factor class (<code>c</code>)</li> <li>Spatial resolution: 30m</li> <li>Begin of time reference: date of first Landsat composite used by the modeling (<code>20240101</code>)</li> <li>End of time reference: date of last Landsat composite used by the modeling (<code>20241231</code>)</li> <li>Spatial extent: global (<code>go</code>)</li> <li>Coordinate system: World Geodetic System 1984, used in GPS (<code>epsg.4326</code>)</li> <li>Version: v2</li> </ol> <h3>Related resources</h3> <ul> <li><strong>Maps of dominant grassland:</strong><br><a href="http://doi.org/10.5281/zenodo.15646181">2000-2002</a><a href="http://doi.org/10.5281/zenodo.15644486"> 2003-2005</a><a href="http://doi.org/10.5281/zenodo.15644623"> 2006-2008</a><a href="http://doi.org/10.5281/zenodo.15647042"> 2009-2011</a><a href="http://doi.org/10.5281/zenodo.15647681"> 2012-2014</a><a href="http://doi.org/10.5281/zenodo.15648306"> 2015-2017</a><a href="http://doi.org/10.5281/zenodo.15648551"> 2018-2020</a><a href="http://doi.org/10.5281/zenodo.15648751"> 2021-2023</a> <a href="http://doi.org/10.5281/zenodo.15649332">2024</a></li> <li><strong>Probability maps of cultivated grassland:</strong><br><a href="https://zenodo.org/records/13890401/files/ggc-30m.csv?download=1">2000-2024 (All URLs)</a></li> <li><strong>Probability maps of natural/semi-natural grassland:</strong><br><a href="https://zenodo.org/records/13890401/files/ggc-30m.csv?download=1">2000-2024 (All URLs)</a></li> <li> <div><strong>Grassland reference samples based on VHR imagery (2000–2024):</strong><br><a href="https://doi.org/10.5281/zenodo.15631655">GeoPackage files</a></div> </li> <li><strong>Global machine learning models (Random Forest):</strong><br><a href="https://doi.org/10.5281/zenodo.13952806">Parquet and joblib python files</a></li> <li><strong>Reference sampling design derived by FSCV:</strong><br><a href="https://doi.org/10.5281/zenodo.11391517">GeoPackage and raster files</a></li> <li><strong>Harmonized reference samples based on existing LULC dataset:</strong><br><a href="https://doi.org/10.5281/zenodo.13951976">GeoPackage and raster files</a></li> <li><strong>Source code for reproducibility:<br></strong><a href="https://doi.org/10.5281/zenodo.13952867">GitHub release</a><strong><br></strong></li> <li><strong>Mapping feedback tool:</strong><br><a href="https://geo-wiki.org">GeoWiki</a></li> <li><strong>Data catalogues:</strong><br><a href="https://stac.openlandmap.org/gpw_ggc-30m/collection.json?.language=en">OpenLandMap STAC</a> <a href="https://global-pasture-watch.projects.earthengine.app/view/ggc-30m">Google Earth Engine</a></li> </ul> <p><strong>Support</strong></p> <p>For questions of bugs/inconsistencies related to the dataset raise a GitHub issue in <a href="https://github.com/wri/global-pasture-watch">https://github.com/wri/global-pasture-watch</a></p>
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.