{
  "fetched_at": "2026-10-19T17:01:00Z",
  "source_url": "https://www.kaggle.com/competitions/esa-eo4soilprotection-2024-predicting-soc-density/data",
  "pages": [
    {
      "name": "data-description",
      "content": "##Training Dataset (train_soc.csv):\n- Dataset Name: KSSL dataset\n- As a part of the SS4GG project, this dataset comprises 5165 unique locations with a total of 23898 laboratory measurements of SOC density.\n##Validation Dataset (test_soc.csv)\nThe dataset contains both public and private test points. Public points are available for internal validation of your models. It will help you fine-tune and assess the accuracy of your predictions during the development phase. Private dataset represents the final test of your model's effectiveness in real-world scenarios. Your goal is to predict these hidden values accurately. \n\n## Columns\nExplanation of columns (variables) in the training dataset can be found [here](https://docs.google.com/document/d/1IbzrJuyOBRFNTzXzcvw_y0O25fVfj2osJFs1TukHflw/edit?usp=sharing).\n\n##Important columns are:\n* `sample_id`: unique ID\n* `logoc_d.f`: target variable in log-scale\n* `X`, `Y`: easting and northing in EPSG 5070\n* `observation_year`: observation year\n* `olc_id`: Open Location ID (https://en.wikipedia.org/wiki/Open_Location_Code) this allows you to check unique locations of points\n* `UUID`: Universally Unique Identifier (https://en.wikipedia.org/wiki/Universally_unique_identifier) this allows you to point to exact sample / unique row; \n\n\nNote that columns `hzn_top` and `hzn_bot` indicate upper and lower boundary of the soil layer (for example 0–30 cm depth interval has upper depth of 0 and lower depth of 30).\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "##Predicting Soil Organic Carbon (SOC) Density:\nSOC is a vital indicator of soil health, affecting fertility and carbon sequestration. In this hackathon, participants are invited to develop innovative models for predicting SOC density in space and time. For this we use a unique 3D+T dataset USDA-NRCS Kellogg Soil Survey Laboratory (KSSL), which provides SOC density measurements at various depth intervals across multiple locations for cca 24,000 samples.\n\n![Figure 1](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F6592816%2F60983c9a9373c35a6812f495225b1c68%2FScreenshot%20from%202024-01-30%2006-50-18.png?generation=1706593840637433&alt=media)",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "The submissions are evaluated using the Root Mean Squared Error (RMSE) between the predicted and observed **log-scaled (log1p) SOC density** of soil samples in the test set:\n$$ RMSE = \\sqrt{(\\frac{1}{n})\\sum_{i=1}^{n}(y_{i} - \\hat{y}_{i})^{2}} $$\n\n\n## Submission File\nFor each `sample_id` in the test set, the participants must predict the SOC content `logoc_d.f` . The file should contain a header and have the following format:\n```\nsample_id,logoc_d.f\n1012,0.54321\n5006,1.54321\n9999,2.54321\n ```",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "##General\n1. Hackathon will run in two rounds. In the first round, top 5 best performing candidates (evaluation criteria in Kaggle) meeting all other requirements will be selected and invited to submit their computational notebooks. In the second round, the best (winner) and 2nd best participants will be selected by the review committee based on their computational notebook. Participants that do not reveal their modeling steps or are not able to reproduce results will not be considered for the best price.\n2. Hackathons / results need to be submitted via Kaggle.com in a timely manner. Deadlines will be followed strictly in second. Submission outside the deadline will not be considered.\n3. Participants will be given a number of covariate layers that they can use to model the target feature (SOC / occurrence of gully erosion), but can also add new layers of choice. Basically participants have complete freedom to choose a modeling framework and add additional covariates, however, results produced need to be reproducible i.e. ALL steps need to be documented in code. The reviewing committee might request to access some additional input data to check that the procedure is correct.\n4. Winners of the hackathons will be invited to present their steps during the event at ESRIN (either online or in person).\n5. Maximum number of participants in the participating team allowed is: 2.\n\n##Submission Limits\nParticipants may submit a maximum of 5 entries per day.\nParticipants may select up to 2 final submissions for judging.\n\n##Competition Timeline\n* Start Date: February 2, 2024\n* End Date: February 16, 2024\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "abstract",
      "content": "Welcome to our hackathon events hosted by the [OpenGeoHub](https://opengeohub.org/) Foundation in collaboration with the European Space Agency (ESA) as part of the **[ESA Symposium on EO 4 Soil Protection and Restoration 6–7 March 2024 at ESA-ESRIN center, Frascati, Italy](https://www.eo4soilprotection.org/)**! The European Space Agency (ESA), along with EO research communities, recognizes the urgency of improving monitoring systems for soil health. These hackathons aim to unite researchers, students, scientists, and technology enthusiasts to address these challenges using open-source remote sensing data and legacy soil data.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Timeline",
      "content": "- Start: 2nd of February, 2024\n- Close: **16th of February 23:59 CET, 2024**\n- Winners announced: 7th March 2024 at the closing session of the Symposium.\n- **The deadlines will be followed strictly.** May the best team win!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": " Rewards and Recognition",
      "content": "Winners of the hackathon will be awarded the chance to showcase their results at ESA’s Symposium on Earth Observation for Soil Protection and Restoration, scheduled in Frascati, Italy, from March 6–7, 2024. Additionally, the winning teams will receive:\n\n- A €15k voucher for cloud processing at ESA’s NoR.\n- A life-long Enhanced Membership for the Open Access PeerJ journals.\n- A 15-minute slot to present their method at the Symposium.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Acknowledgements",
      "content": " These hackathons were made possible thanks to the AI4SoilHealth and Open-Earth-Monitor Cyberinfrastructure projects. [AI4SoilHealth](https://ai4soilhealth.eu/) project has received funding from the European Union’s Horizon Europe research and innovation programme under grant agreement [No. 101086179](https://cordis.europa.eu/project/id/101086179). [EarthMonitor.org](https://earthmonitor.org/) project has received funding from the European Union’s Horizon Europe research an innovation programme under grant agreement [No. 101059548](https://cordis.europa.eu/project/id/101059548). \n\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Literature of interest",
      "content": "- Bouasria, A., Bouslihim, Y., Gupta, S., Taghizadeh-Mehrjardi, R., & Hengl, T. (2023). Predictive performance of machine learning model with varying sampling designs, sample sizes, and spatial extents. Ecological Informatics, 78, 102294. https://doi.org/10.1016/j.ecoinf.2023.102294 \n- Hengl, T., Sorenson, P., Parente, L., Cornish, K., Battigelli, J., Bonannella, C., ... & Nichols, K. (2023). Assessment of soil organic carbon stocks in Alberta using 2-scale sampling and 3D predictive soil mapping. FACETS, 8, 1-17. https://doi.org/10.1139/facets-2023-0040 \n- Maxwell, T. L., Hengl, T., Parente, L. L., Minarik, R., Worthington, T. A., Bunting, P., ... & Landis, E. (2023). Global mangrove soil organic carbon stocks dataset at 30 m resolution for the year 2020 based on spatiotemporal predictive machine learning. Data in Brief, 50, 109621. https://doi.org/10.1016/j.dib.2023.109621 \n- Sanderman, J., Savage, K., & Dangal, S. R. (2020). Mid‐infrared spectroscopy for prediction of soil health indicators in the United States. Soil Science Society of America Journal, 84(1), 251-261. https://doi.org/10.1002/saj2.20009 \n- Ugbaje, S. U., Karunaratne, S., Bishop, T., Gregory, L., Searle, R., Coelli, K., & Farrell, M. (2024). Space-time mapping of soil organic carbon stock and its local drivers: Potential for use in carbon accounting. Geoderma, 441, 116771. https://doi.org/10.1016/j.geoderma.2023.116771 \n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}