Human Modification Index forecasts from 2025 to 2040

Global probabilistic forecasts of the Human Modification (HM) index at 1 km resolution and 5-year intervals from 2025 to 2040. Forecasts were produced by a convolutional LSTM trained on the harmonized 1990–2020 HM time series of Theobald et al. (2025), together with dynamic covariates (HM component stressors, gridded population, gridded GDP) and static covariates (CHELSA climate, EarthEnv elevation, WDPA protected-area masks, and a development suitability index). The model produces independent predictions at four horizons (2025, 2030, 2035, 2040), each accompanied by a central estimate and calibrated 2.5th and 97.5th percentile bounds that together span a 95% prediction interval.The repository contains 12 single-band float32 GeoTIFFs in EPSG:4326 covering land between 84°N and 70°S, named hm_QUANTILE_YEAR.tiff (YEAR ∈ {2025, 2030, 2035, 2040}; QUANTILE ∈ {lower, central, upper}). Pixel values range from 0 (unmodified) to 1 (completely modified). For convenience, the historical HM observations from 1990–2020, resampled and masked to match the forecast grid, are also included as hm_observed_YEAR.tiff, so users can work with a seamless 1990–2040 time series.