Find research datasets worth reusing

Modelling environmental suitability of sorghum, wheat and maize in Europe under climate change (Code & data)

Wheat and maize play an important role as crops for human consumption and animal feed in Europe. To guarantee food security and the stability of the agricultural sector in Europe, it is crucial to determine how climate change will impact the environmental suitability and thus the potential geographic distribution of these crops. Sorghum, a crop that originates in Africa, has seen a recent increase in cultivation in Europe. Due to its tolerance to more extreme climate conditions and its versatility of use, it might inherit a high potential as an alternative crop. Occurrence data of sorghum, wheat and maize as well as several environmental variables were used as input data for an ensemble modelling approach that averages machine learning models for species distribution modelling (SDM). CHELSA served as a source for present bioclimatic conditions and future climate scenarios, namely SSP126 and SSP370 for the period 2041-2070, and HSWD supplied soil variables, since both climate and soil influence crop development. A set of models was evaluated to select the best performing models for the ensemble modelling. The ensemble models were extrapolated to the future scenarios to predict geographic shifts of suitable cultivation areas due to climate change and analyze sorghum’s potential as an alternative crop. Under the climate scenarios, the three crops saw a shift of suitability in Europe with losses in Southern Europe and expansions of suitable environmental conditions in the northeast of Europe. Sorghum was the crop with the highest potential to replace maize and wheat in Southern Europe in areas where they lose suitability under climate change. Therefore, sorghum confirmed its function as an alternative crop. It also was the crop that benefits consistently from climate change, growing its total suitable area in Europe under both climate scenarios. Maize loses total suitable area in Europe in both climate scenarios but kept the highest amount of total suitable area in Europe in all projected time periods. The outcome of this study is of high importance for European farmers and policy makers as it enables them to apply effective adaptation and mitigation strategies that will support crop production under future climate conditions.

SDUST2024MSS_AO: a mean sea surface model of the Arctic Ocean based on CryoSat-2 SAR altimeter data

This model is a mean sea surface model for ice-covered regions, using CryoSat-2 satellite SAR mode altimeter data from July 2010 to December 2023. The heights are referenced to the WGS-84 ellipsoid, and the grid size is 5 km × 5 km.

Modelled temperature, mortality impact, and external benefits of cool roofs and rooftop photovoltaics in London - supporting data

Supporting data for "Modelled temperature, mortality impact, and external benefits of cool roofs and rooftop photovoltaics in London" Included are outputs from the Weather Research and Forecast (WRF) model. All simulations cover London, United Kingdom over summer 2018. Scenarios include a "baseline" which represents the real urban climate of the region, and scenarios which model 100% coverage of rooftops with either high albedo materials or solar panels. Data are provided in netCDF format. <ul> <li>The baseline simulation which models the current urban climate of the region WRF_Urb_BouLac_T2-V10-U10_20180525-20180831.nc</li> <li>The 100% rooftop-solar simulation WRF_Urb_BouLac_PV_T2-V10-U10-PSFC-RAINNC-TH2-Q2_20180525-20180831.nc</li> <li>The 100% high-albedo roof simulation WRF_Urb_BouLac_ClRf_T2-V10-U10-PSFC-RAINNC-TH2-Q2_20180525-20180831.nc</li> <li>The non-urban scenario is in WRF_NoUrb_BouLac_T2-V10-U10-PSFC-RAINNC-TH2-Q2_20180525-20180831.nc</li> <li>The power production estimates solarpv_prod_2018.nc</li> <li>wrf_popweighting-main.zip contains the analysis code. It is provided as-is with no guarantee of usability.</li> </ul> T2 means air temperature at 2m height. V10 and U10 are windspeeds at 10m height. PSFC is surface level pressure. TH2 is potential temperature. Q2 is specific humidity at 2m height. More description of the simulations is given in the citing article.

Secondary Data for: Enhanced Modeling of Back-Mixing in Chemical Reactor Networks

Secondary data for the results presented in the preprint "Enhanced Modeling of Back-Mixing in Chemical Reactor Networks" by L. Gossel, M. Fricke and D. Bothe (2023).  https://arxiv.org/abs/2305.11591 Tables containing the secondary data of the results presented in Figure 5, a-d are provided.  The used code is confidential and thus not included in the repository.  Funded by the Hessian Ministry of Higher Education, Research, Science and the Arts - cluster project Clean Circles.

opencc-by-4.0Mar 2024View details →

Exploring deep learning models for 4D-STEM-DPC data processing

This repository contains scanning transmission electron microscopy data and processing files used in the journal publication "Exploring deep learning models for 4D-STEM-DPC data processing". DOI: <a href="https://doi.org/10.1016/j.ultramic.2024.114058">10.1016/j.ultramic.2024.114058</a> Prerequisites The scripts presented below require certain open-source Python packages to run. Library versions used to run the scripts are: <ul> <li>hyperspy 1.7.1</li> <li>pyxem 0.14.2</li> <li>fpd 0.2.5</li> <li>pytorch 1.12.1 (cudatoolkit 11.6.0)</li> <li>jupyterlab 4.0.7</li> </ul> Data files Three zipped folders are included. Two of them contain the training- and inference data for the neural networks, aptly named training_data.zip and inference_data.zip. PyTorch state dictionaries for trained models are included in the models.zip folder. Processing scripts All scripts are included in an IPython notebook format (.ipynb extension). The notebooks Segmentation.ipynb and Regression.ipynb contain the code for training and inference of the segmentation and regression models, respectively. The Training_data_creation.ipynb notebook contains the code to preprocess the training data for both neural network models. The Standard_algorithms.ipynb notebook has the code for doing center of mass and edge filtering/disc detection algorithms for STEM-DPC processing.

Runoff from Greenland's Firn Area - Why do MODIS, RCMs and a Firn Model disagree? - Code and Data

<h1>README – Overview code and data for obtaining MODIS runoff limits and comparison to and between MAR and RACMO / IMAU-FDM</h1>   <h2>Basic workflow</h2> Table 1 lists the projects, the pieces of code in the projects and the key output of each piece of code. This serves at illustrating the basic workflow. More detailed information on required input is provided below. The README files of the various projects provide details on how to use the code.   Table 1: This table might not display properly, please refer to the file _README_overview.pdf. General project overview. The projects and code are shown in the order they are intended to be used. In italic are parts of code that were not used for Machguth et al. (in review) but instead for Machguth et al. (2022). They might not be fully compatible anymore with runoff limits Yr calculated along flowline-polygons. <table> <tbody> <tr> <td> Project </td> <td> Code </td> <td> Output </td> </tr> <tr> <td> flowlines </td> <td> crop_gdalwarp.py </td> <td> DEM, cropped and reprojected to required size and grid (identical to the MODIS files) </td> </tr> <tr> <td> flowline_seedpoints.py </td> <td> Seedpoints to calculate flowlines </td> </tr> <tr> <td> flowline.py </td> <td> Flowlines and flowline- polygons </td> </tr> <tr> <td> MODIS_Greenland </td> <td> MODIS_array_filter.py </td> <td> Daily MODIS grids filtered for outliers </td> </tr> <tr> <td> MODIS_stddev_spatial.py </td> <td> Daily grids of MODIS spatial standard deviation </td> </tr> <tr> <td> MODIS_mean_stddev.py </td> <td> Greenland-wide map of background MODIS spatial standard deviation </td> </tr> <tr> <td> MODIS_NDWI_tiff_to_nc.py </td> <td> MODIS NDWI converted from tiff to netCDF </td> </tr> <tr> <td> MODIS_find_slush_limit.py </td> <td> Table of all detected daily Yr </td> </tr> <tr> <td> Greenland_RCM_analysis_prep </td> <td> modis_vs_mar.py </td> <td> Tables of daily MAR Yr for all flowline-polygons </td> </tr> <tr> <td> modis_vs_racmo.py </td> <td> Table of daily RACMO Yr for all flowline-polygons </td> </tr> <tr> <td> MODIS_Greenland_analysis </td> <td> MODIS_A_SLmax.py </td> <td> Calculates annual maxima of the Yr, written into table. Creates plots visualizing MODIS Yr spatial and temporal distribution – only used in Machguth et al. (2022) </td> </tr> <tr> <td> MODIS_A_SLanalyis.py </td> <td> Plots that visualize progression and forcing behind MODIS Yr – only used in Machguth et al. (2022) </td> </tr> <tr> <td> MODIS_comp_RCM_maxYs.py </td> <td> For all of Greenland: Plots that compare RCM and MODIS maxYr </td> </tr> <tr> <td> MODIS_comp_RCM_Ys.py </td> <td> For all of Greenland: Plots that compare RCM and MODIS daily Yr </td> </tr> <tr> <td> Greenland_RCM_analysis </td> <td> RCM_analysis_comp_Ktransect.py </td> <td> For the K-transect: Plots/tables that compare RACMO/FDM and MAR Yr as well as various RCM parameters </td> </tr> <tr> <td> Jupyter notebooks </td> <td> K-transect_MAR_Greenland-wide_to_K-transect.ipynb </td> <td> Writes MAR .nc files that contain data only for the K-Transect </td> </tr> <tr> <td> K-transect_RACMO-MAR_depth-data-plot.ipynb </td> <td> For the K-transect: Plots to compare depth information of MAR and RACMO/FDM </td> </tr> </tbody> </table>     <h2>Input and output data</h2> Unless mentioned otherwise, all input data required is provided in this repository. Input files for one piece of code are at the same time the output of a piece of code that needs to be run before. Note 1: due to space constraints only the so called “test” data set for runoff limit detection is provided. The full input data to Greenland-wide runoff limit detection are too large in volume but can be obtained from the authors. Note 2: While we only provide the test data for the MODIS runoff limit detection, we provide the full output of the Greenland-wide runoff limit detection (these data are small in volume). <h3>Input to flowlines/crop_gdalwarp.py</h3> arcticdem_mosaic_100m_v30_greenland_icesheet_geoidCorr.tif <h3> </h3> <h3>Input to flowlines/flowline_seedpoints.py</h3> seedline_v2.1.shp <h3> </h3> <h3>Input to flowlines/flowline.py</h3> è In the case of using the test data, no seed file exists. Do not provide one, seed points will be calculated along a north-south line as specified. greenland_vel_mosaic200_2015-2018_vx_v02-composite-crop.tif greenland_vel_mosaic200_2015-2018_vy_v02-composite-crop.tif dem_test_gapfilled.tif or for Greenland-wide: greenland_vel_mosaic500_2015-2018_vx_v02-composite-crop.tif greenland_vel_mosaic500_2015-2018_vy_v02-composite-crop.tif arcticdem_mosaic_500m_v30_greenland_icesheet_geoidCorr_GapFilled.tif seedpoints_v3.4.shp <h3> </h3> <h3>Input to MODIS_Greenland/MODIS_array_filter.py</h3> /sat_modis_proc_GR_l1test/*   <h3>Input to MODIS_Greenland/MODIS_stddev_spatial.py</h3> /sat_modis_proc_test_l2/*   <h3>Input to MODIS_Greenland/MODIS_mean_stddev.py</h3> /sat_modis_proc_test_l3/* dem_test_gapfilled.tif   <h3>Input to MODIS_Greenland/ MODIS_NDWI_tiff_to_nc.py</h3> è Input data were computed directly on Google Earth Engine, they have not been preserved, only output data exist   <h3>Input to MODIS_Greenland/MODIS_find_slush_limit.py</h3> /sat_modis_proc_test_l3/* /sat_modis_proc_test_l2_NDWI/* mask_greenland_icesheet/dem_test_gapfilled.tif Ys_polygons__test_W20km.shp flowlines__test_W20km.shp test_MOD10A1.l3.v4_yrs2000-2021_doy126-136_stddev_median.tif   <h3>Input to Greenland_RCM_analysis_prep/modis_vs_mar.py</h3> /flash/tedstona/MARv.HorstRCMStudy_20240624/* what is the DEM, probably the same as used elsewhere? flowline Polygons /flash/tedstona/_list_PolyIDs.xlsx Why output still to 'MAR-v3.12.1-rlim-slush.nc'? Probably simply not changed without any effect? <h3> </h3> <h3>Input to Greenland_RCM_analysis_prep/modis_vs_racmo.py</h3> /flash/tedstona/RACMO/1km/runoff/* arcticdem_mosaic_500m_v30_greenland_icesheet_GeoidCorr_GapFilled_RACMO1km.tif racmo_polys.nc   <h3>Input to MODIS_Greenland_analysis/MODIS_A_SLmax.py</h3> _test_slush-limit_output_table.xlsx or the Greenland-wide output data: _GR_slush-limit_output_table.xlsx   <h3>Input to MODIS_Greenland_analysis/MODIS_A_SLanalyis.py</h3> KAN_U_hourly_v3_fewer_columns.xlsx KAN_M_hourly_v3_fewer_columns.xlsx (selected_SL_years_and_stripes_20km.xlsx: not found and was also only used in Machguth et al., 2022) __test_table_complete_annual_max_SL.xlsx __test_slush-limit_output_table_OnlyValidEntries.xlsx or the Greenland-wide output data: __GR_table_complete_annual_max_SL.xlsx __GR_test_slush-limit_output_table_OnlyValidEntries.xlsx   <h3>Input to MODIS_Greenland_analysis/MODIS_comp_RCM_maxYs.py</h3> è This file has not been tested whether it also works with the “test” data RACMO2.3p2_ERA5_3h_FGRN055.1km-rlim-RUa1mm.xlsx MAR-v.20240624-rlim-RUa1mm.xlsx _stripes_with_aquifers.xlsx __test_table_simple_annual_max_SL.xlsx __test_table_complete_annual_max_SL.xlsx or the Greenland-wide output data: __GR_table_simple_annual_max_SL.xlsx __GR_table_complete_annual_max_SL.xlsx   <h3>Input to MODIS_Greenland_analysis/MODIS_comp_RCM_Ys.py</h3> è This file has not been tested whether it also works with the “test” data __GR_slush-limit_output_table_OnlyValidEntries.xlsx flowlines_daily_rlims_RACMO_1mmEvents_10mmAnnual_2000_2021.xlsx flowlines_daily_rlims_MAR-v.20240624_1mmEvents_10mmAnnual_2000_2021.xlsx   <h3>Input to Greenland_RCM_analysis/RCM_analysis_comp_Ktransect.py</h3> è Certain data sets (indicated below) are not provided as they would exceed the available space in the repository. They can be obtained at no conditions from the authors è This file has not been tested whether it would also works with the “test” data Too large: /FDM_Greenland-wide/* Too large: /RACMO-FDM_K-Transect_updated/* Too large: /RACMO_Greenland-wide/* Too large: /MARv3.14_K-transect_continuous/* Too large: FGRN055_Masks.nc __GR_slush-limit_output_table_OnlyValidEntries.xlsx __GR_table_simple_annual_max_SL.xlsx __GR_table_complete_annual_max_SL.xlsx   <h3>Input to _notebooks\K-transect_MAR_Greenland-wide_to_K-transect.ipynb</h3>   <h3>Input to _notebooks\K-transect_RACMO-MAR_depth-data-plot.ipynb</h3>   <h2>Output data</h2> The basic output files of the runoff limit detection for all of Greenland are: _GR_slush-limit_output_table.xlsx (written by MODIS_Greenland/MODIS_find_slush_limit.py) __GR_slush-limit_output_table_OnlyValidEntries.xlsx (by MODIS_Greenland_analysis/MODIS_A_SLmax.py) __GR_table_simple_annual_max_SL.xlsx  (by MODIS_Greenland_analysis/MODIS_A_SLmax.py) __GR_table_complete_annual_max_SL.xlsx  (by MODIS_Greenland_analysis/MODIS_A_SLmax.py)   The basic output files of the runoff limit detection for the “test” data set are: _test_slush-limit_output_table.xlsx (written by MODIS_Greenland/MODIS_find_slush_limit.py) __test_slush-limit_output_table_OnlyValidEntries.xlsx (by MODIS_Greenland_analysis/MODIS_A_SLmax.py) __test_table_simple_annual_max_SL.xlsx  (by MODIS_Greenland_analysis/MODIS_A_SLmax.py) __test_table_complete_annual_max_SL.xlsx  (by MODIS_Greenland_analysis/MODIS_A_SLmax.py)

Sample data for "Classification Modeling for Hazardous Rip Current Prediction" Notebook

This sample dataset is used in the notebook "Classification Modeling for Hazardous Rip Current Prediction" to demonstrate the application of using machine learning to identify hazardous rip current.  The notebook is available in the NOAA Center for Artificial Intelligence GitHub Learning Journey repository (https://github.com/noaa-ncai/learning-journey). The full dataset is available via NOAA.

Raw data to "Order-by-disorder in the antiferromagnetic $J_1$-$J_2$-$J_3$ transverse-field Ising model on the ruby lattice"

This directory contains the data used to generate the results in the work "Order-by-disorder in the antiferromagnetic $J_1$-$J_2$-$J_3$ transverse-field Ising model on the ruby lattice" [1]. To get an overview of the organization of the directory and a description of the data we recommend the README.txt file. [1]: A. Duft et al., Order-by-disorder in the antiferromagnetic $J_1$-$J_2$-$J_3$ transverse-field Ising model on the ruby lattice. arXiv:2312.12941

opencc-by-4.0Jul 2022View details →

Topic Modeling Data

This repository contains the data used to perform topic modeling on premodern texts; the article that discusses the application and the outcomes can be found here: https://doi.org/10.57813/20220623-153139-0

Data and models for: Learning Ordering in Crystalline Materials with Symmetry-Aware Graph Neural Networks

Data (ver 1.1) and trained models for our paper "<a href="https://arxiv.org/abs/2409.13851">Learning Ordering in Crystalline Materials with Symmetry-Aware Graph Neural Networks</a>". If you use such data or models, please cite our paper. These three directories need to be downloaded and copied into our source codes in order to reproduce our paper: <a href="https://github.com/learningmatter-mit/PerovskiteOrderingGCNNs">https://github.com/learningmatter-mit/PerovskiteOrderingGCNNs</a> <ul> <li>data: All data files for training and evaluating GCNNs, with a copy archived on the Materials Data Facility (<a href="https://doi.org/10.18126/ncqt-rh18">DOI: 10.18126/ncqt-rh18</a>)</li> <li>saved_models: All saved model files for evaluating GCNNs</li> <li>best_models: All best model files for evaluating GCNNs</li> </ul>

Dataset and Data Dictionary for "Enhancing Consumer Satisfaction in Live Commerce: A Study of Middle-Aged Women's Cosmetics Purchases Using TAM, PVT, and SIT Models

This dataset is part of a study investigating the underexplored factors driving middle-aged Chinese women’s purchasing behavior in live commerce, particularly in the context of their decision-making amidst the rapid expansion of e-commerce. The study employs a comprehensive theoretical framework based on the Technology Acceptance Model (TAM), Perceived Value Theory (PVT), and Social Influence Theory (SIT). Data were collected through a structured survey administered to 653 women aged 40 to 59. The dataset captures key variables including ease of use, pricing, consumer trust, platform interactivity, and purchase satisfaction. These variables are essential for understanding the complex relationships that influence purchasing decisions in live-stream shopping environments. The dataset has been analyzed using Structural Equation Modeling (SEM), revealing that factors such as ease of use, perceived value from competitive pricing, consumer trust, and real-time platform interactivity significantly enhance purchase satisfaction. Moreover, the results demonstrate that perceived value moderates these relationships, amplifying their effects under conditions of high perceived value. This dataset provides valuable insights into the psychological and social factors that shape e-commerce behavior, offering implications for optimizing platform design to promote consumer trust and long-term engagement. Keywords: Live commerce, Middle-aged women, Purchasing behavior, Technology Acceptance Model (TAM), Perceived Value Theory (PVT), Social Influence Theory (SIT), Structural Equation Modeling (SEM), Consumer trust, E-commerce.

model data for "Observed and Model-Simulated Dramatic Bottom Temperature Variations During a Weakened Typhoon in the Northern Yellow Sea"-part06

This dataset is part six of the FVCOM model data of typhoon Lekima during August 2019. The data is in netcdf format.  Each nc file contained varaibles including sea level elevation, current, temperature, salinity and mixing parameters of 24 hours dring one day.

model data for "Observed and Model-Simulated Dramatic Bottom Temperature Variations During a Weakened Typhoon in the Northern Yellow Sea"-part05

This dataset is part five of the FVCOM model data of typhoon Lekima during August 2019. The data is in netcdf format.  Each nc file contained varaibles including sea level elevation, current, temperature, salinity and mixing parameters of 24 hours dring one day.

model data for "Observed and Model-Simulated Dramatic Bottom Temperature Variations During a Weakened Typhoon in the Northern Yellow Sea"-part02

This dataset is part two of the FVCOM model data of typhoon Lekima during August 2019. The data is in netcdf format.  Each nc file contained varaibles including sea level elevation, current, temperature, salinity and mixing parameters of 24 hours dring one day.

model data for "Observed and Model-Simulated Dramatic Bottom Temperature Variations During a Weakened Typhoon in the Northern Yellow Sea"-part03

This dataset is part three of the FVCOM model data of typhoon Lekima during August 2019. The data is in netcdf format.  Each nc file contained varaibles including sea level elevation, current, temperature, salinity and mixing parameters of 24 hours dring one day.

model data for "Observed and Model-Simulated Dramatic Bottom Temperature Variations During a Weakened Typhoon in the Northern Yellow Sea"-part04

This dataset is part one of the? FVCOM model data of typhoon Lekima during August 2019. The data is in netcdf format.  Each nc file contained varaibles including sea level elevation, current, temperature, salinity and mixing parameters of 24 hours dring one day.

model data for "Observed and Model-Simulated Dramatic Bottom Temperature Variations During a Weakened Typhoon in the Northern Yellow Sea"-Part01

This dataset is part one of the  FVCOM model data of typhoon Lekima during August 2019. The data is in netcdf format. Each nc file contained varaibles including sea level elevation, current, temperature, salinity and mixing parameters of 24 hours dring one day.

Data for: Accurate state-of-charge estimation for sodium-ion batteries based on a low-complexity model with hierarchical learning

The dataset accompanies the Journal of Energy Storage publication by Shuquan Wang et al. (2024), Accurate state-of-charge estimation for sodium-ion batteries based on a low-complexity model with hierarchical learning, DOI 10.1016/j.est.2024.112571.  <h2>Experimental Description:</h2> The dataset comprises results from two experimental tests: pulse testing and driving cycle testing. These tests were conducted on two types of sodium-ion batteries—one with a capacity of 3.2 Ah (battery numbers: 1, 2, and 5) and another with a capacity of 10 Ah (battery numbers: 3, 4, and 6). <h3>Pulse Testing:</h3> The pulse tests were carried out using a battery test platform, consisting of an Arbin battery testing system, a temperature-controlled chamber, and a computer. The tests were performed on two 3.2 Ah and two 10 Ah sodium-ion batteries from Transimage and HiNa, respectively, with a nominal voltage of 3.0 V. The upper and lower cut-off voltages were set at 3.9 V and 1.5 V. Enhanced pulse tests were conducted at six different temperatures: -5 ℃, 5 °C, 15 ℃, 25 ℃, 35 ℃, and 45 ℃. The state-of-charge (SOC) was varied in 10% intervals, with pulse currents escalating incrementally from 0.25C to 3C at 0.25C intervals. Each pulse lasted for 5 seconds, followed by a 15-second rest. After completing each set of pulses, the current was increased, and the process was repeated with a two-minute pause between sets of pulses. <h3>Driving Cycle Testing:</h3> The driving cycle tests were designed to simulate real-world driving conditions using various standard test methods, including the Federal Urban Driving Schedule (FUDS), Urban Dynamometer Driving Schedule (UDDS), and Dynamic Stress Test (DST). These tests were performed in a temperature-controlled chamber using both the 3.2 Ah and 10 Ah sodium-ion batteries. As with the pulse tests, driving cycle tests were carried out at temperatures of -5 ℃, 5 °C, 15 ℃, 25 ℃, 35 ℃, and 45 ℃. Before each test, the batteries were charged with a 0.5C constant current-constant voltage (CC-CV) charging protocol up to 3.9 V, with a cut-off current of 0.02C. After a 30-minute rest, the driving cycle protocol was performed for seven iterations. <h2>File Naming Conventions:</h2> The dataset files are named based on the experimental conditions, as follows: <ul> <li>Pulse_data_tempX_batY: Data from the pulse tests, where X represents the testing temperature and Y denotes the battery number.</li> <li>Driving_cycle_data_tempX_batY: Data from the driving cycle tests, where X represents the testing temperature and Y denotes the battery number.</li> </ul>

opencc-by-4.0Jul 2024View details →

Modeling of Greenland Peripheral Glaciers - Supporting data

This data is related to section 4.9 (Figure 12) in the "Kanzow, T., Humbert, A., Mölg, T., Scheinert, M., Braun, M., Burchard, H., Doglioni, F., Hochreuther, P., Horwath, M., Huhn, O., Kusche, J., Loebel, E., Lutz, K., Marzeion, B., McPherson, R., Mohammadi-Aragh, M., Möller, M., Pickler, C., Reinert, M., Rhein, M., Rückamp, M., Schaffer, J., Shafeeque, M., Stolzenberger, S., Timmermann, R., Turton, J., Wekerle, C., and Zeising, O.: The atmosphere-land/ice-ocean system in the region near the 79N Glacier in Northeast Greenland: Synthesis and key findings from GROCE, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2024-757, 2024."

OEMC Hackathon 2023: Global FAPAR Modeling Dataset (including raster data)

Dataset organized by the <a href="https://earthmonitor.org/">Open-Earth-Monitor (OEMC) project</a> within the context of <a href="http://www.kaggle.com/competitions/oemc-hackathon-eu-land-cover-classification/overview">Hackathon 2023</a>. The dataset contains monthly mean FAPAR values aggregated by each ground station. FAPAR represents the fraction of the incoming (photosynthetic active) radiation that is absorbed by vegetation, and is given in the range <code>0-1</code>. It is a measure of vegetation health and ecosystem functioning, and a key parameter in light use efficiency models that model primary productivity. For each monthly FAPAR value, a set of covariates / features were extracted from 32 raster spatial layers, including including satellite (spectral bands and indices) and temperature images (land surface temperature), climate images (precipitation) and digital terrain model (slope and elevation). The features are organized by columns, unique data points in time are identified by the <code>sample_id</code> column, and data points points belonging to the same location are identified by <code>station_number</code>. Column names: <ul> <li><code>sample_id</code>: unique identifier of datapoint</li> <li><code>station</code>: ground station number</li> <li><code>fapar</code>: monthly mean FAPAR</li> <li><code>month</code>: month of measurement</li> <li><code>modis_{..}</code>: NDVI, EVI, reflectance bands 1 (red), 2 (near-infrared), 3 (blue), and 7 (mid-infrared) based on <a href="https://lpdaac.usgs.gov/products/mod13q1v061/">MOD13Q1</a></li> <li><code>modis_lst_day_p{..}</code>: Land surface temperatures daytime of percentiles 5th, 50th and 95th based on <a href="https://lpdaac.usgs.gov/products/mod11a2v061/">MOD11A2</a></li> <li><code>modis_lst_night_p{..}</code>: Land surface temperatures nighttime of percentiles 5th, 50th and 95th based on <a href="https://lpdaac.usgs.gov/products/mod11a2v061/">MOD11A2</a></li> <li><code>wv_yearly_p{..}</code>: Water vapour aggregated yearly by percentiles 25th, 50th and 75th based on derived from <a href="https://zenodo.org/record/8226282">MCD19A2</a></li> <li><code>wv_monthly_lt_p{..}</code>: Water vapour aggregated long-term monthly by percentiles 25th, 50th and 75th based on <a href="https://zenodo.org/record/8226282">MCD19A2</a></li> <li><code>wv_monthly_lt_sd</code>: Water vapour aggregated long-term monthly standard deviation based on <a href="https://zenodo.org/record/8226282">MCD19A2</a></li> <li><code>wv_monthly_ts_raw</code>: Water vapour monthly time series based on <a href="https://zenodo.org/record/8226282">MCD19A2</a></li> <li><code>wv_monthly_ts_smooth</code>: Water vapour monthly time series smoothed using the Whittaker method based on <a href="https://zenodo.org/record/8226282">MCD19A2</a></li> <li><code>accum_pr_monthly</code>: Monthly accumulated precipitation based on <a href="https://doi.org/10.1038/sdata.2017.122">CHELSA timeseries</a></li> <li><code>dtm_{..}</code>: Several DTM derivatives (Elevation, Slope, aspect (sine, cosine), curvature (up- and downslope), openness (negative, positive), compound topographic index (cti), valley bottom flatness (vbf)) based on <a href="https://ui.adsabs.harvard.edu/abs/2017AGUFM.H12C..04Y">MERIT DEM</a></li> </ul> Files <ul> <li>train.csv: Training set with 3,461 rows and 36 columns, including sample id (<code>sample_id</code> - index column), ground station (<code>station</code>), reference month (<code>month</code>), measured FAPAR (<code>fapar</code>), and 32 features / covariates</li> <li>test.csv: Test set with 4,939 rows and 34 columns, including sample id (<code>sample_id</code> - index column), ground station (<code>station</code>), reference month (<code>month</code>) and 32 features / covariates</li> <li>sample_submission.csv: a sample submission file with 4,939 rows and 2 columns, including sample id (<code>sample_id</code> - index column) and measured FAPAR (<code>fapar</code>)</li> </ul>