Data for: A fundamental trade-off among resilience, resistance, redundancy, and efficiency across the biosphere

Gross primary productivity (GPP) was extracted from the widely-available NASA MOD17 MODIS GPP product for the dates from 7/4/2001 to 2/26/2022, at 500 m resolution and at 8-day time steps referred to as “epochs”. The NASA MOD17A2H product is a function of vegetative leaf area coverage (via satellite imagery), incident photosynthetically active radiation (via meteorological records), surface air temperature (via meteorological records), vapor pressure deficit (via meteorological records), and light use efficiency (via algorithms developed from eddy covariance flux tower measurements across dominant vegetation types). GPP was extracted from MOD17 and converted into units of grams of carbon per meter squared per day. For more information on MOD17, we refer the reader to NASA’s extensive literature on the product (1–3). We sought to create a high quality gross primary production (GPP) dataset representative of the 3.8 billion MOD17 satellite pixels that span the globe (at 500 m resolution). We delineated the terrestrial world into 14 biomes using the World Wildlife Fund For Nature’s (WWF) world land cover map, where each biome possessed characteristic vegetation community types adapted to the specific climate and topographic relief. We then identified 2,616,334 georeferenced locations that passed several QA/QC filters. We observed low ecoregion representation in several biomes due to water and cloud contamination, and therefore to improve spatial coverage and biome representation in our final analysis, we allowed partial clouds and used a 50% and greater QA filter for the following six biomes: tropical and subtropical moist broadleaf forests; mangrove forests; temperate broadleaf and mixed forests; temperate grasslands, savannas, and shrublands; tundra; and taiga/boreal forests. We then randomly sampled from among these locations in proportion to the land area of each biome, yielding a final GPP dataset for the biosphere containing 777,656 locations. Our initial biosphere-wide analysis included all 14 WWF delineated biomes, however in our final analysis, tundra and boreal forest/taiga biomes were excluded because of concerns regarding the QA/QC uncertainty of MODIS GPP estimates in these biomes due to snow and ice coverage during a short and rapidly progressing growing season, high cloud frequency, and low sun-angles. Due to these concerns about the consistency and reliability of the data in these high-latitude biomes compared to the other twelve biomes, locations from tundra and taiga/boreal forest biomes were removed from the biosphere-wide dataset (BiosphereGPP_noTaigaTundra.csv), and were not included in our final analysis. GPP for these two biomes are included despite the noted limitations with their data reliability (tun50.csv and tai50.csv). The following 12 WWF biome classes were represented in our final analysis: 1) tropical and subtropical moist broadleaf forests, 2) tropical and subtropical dry broadleaf forests, 3) tropical and subtropical coniferous forests, 4) temperate broadleaf and mixed forests, 5) temperate coniferous forests, 6) tropical and subtropical grasslands, savannas, and shrublands, 7) temperate grasslands, savannas, and shrublands, 8) flooded grasslands and savannas, 9) montane grasslands and shrublands, 10) Mediterranean forests, woodlands, and scrub, 11) deserts and xeric shrublands, and 12) mangrove forests. REFERENCES: (1) S. W. Running, M. Zhao, Daily GPP and annual NPP (MOD17A2/A3) products NASA Earth Observing System MODIS land algorithm. MOD17 User’s Guide 2015, 1-28 (2015). (2) F. A. Heinsch et al., GPP and NPP (mod17a2/a3) products nasa modis land algorithm. MOD17 user’s guide, 1-57 (2003). (3) S. W. Running, R. Nemani, J. M. Glassy, P. E. Thornton (1999) MODIS daily photosynthesis (PSN) and annual net primary production (NPP) product (MOD17) Algorithm Theoretical Basis Document.