Spatial-temporal patterns and environmental controls of soil organic carbon stocks in global tidal wetlands since 2009

Dataset and Code for "Spatial-temporal patterns and environmental controls of soil organic carbon stocks in global tidal wetlands since 2009"

Abstract: Tidal wetlands are important blue carbon ecosystems with high carbon storage, making them a global research priority for climate mitigation. However, current global estimates of tidal wetland soil organic carbon (SOC) stocks mainly focus on individual wetland types and infer SOC stock changes based on wetland area changes, leaving our understanding of coastal wetland carbon stocks, dynamics, and cross-wetland type mechanisms unclear. By jointly accounting for changes in both SOC density and wetland extent, this study provides the first spatially explicit and temporally resolved global assessment of SOC stocks and decadal changes across mangroves, salt marshes, and tidal flats. Using a spatio-temporal digital soil mapping framework with 5,489 soil observations (42,154 soil layer records) and 63 environmental covariates, we generated 90 m SOC maps at three depths (0 ~ 30 cm, 30 ~ 60 cm, and 60 ~ 100 cm) for two periods (2009-2014 and 2015-2020). The predicted global SOC stocks in mangroves (mean: 4.85 Pg), salt marshes (0.85 Pg), and tidal flats (2.57 Pg) account for 58.6%, 10.2%, and 31.1% of the total global SOC stocks in 2015-2020, respectively.  Results reveal a 6.11% decline in global total, dropping from 8.80 Pg to 8.27 Pg, with greater losses in deeper layers. Low-latitude regions such as South Asia and Africa were key loss hotspots. SOC losses were more pronounced in lower-middle and upper-middle income countries. Mangroves suffered the largest losses among the three, primarily due to SOC density decline, while salt marshes gained in both area and SOC density. Climate, elevation, and total suspended matter drove the spatiotemporal SOC in mangroves, whereas oceanographic factors and depth influenced SOC in salt marshes. In tidal flats, oceanographic factors played the dominant role. This work reveals the magnitude and drivers of blue carbon change across different wetland types, underscoring the need for targeted conservation to support sustainable management and climate mitigation.

File Description: 

Each .7z file contains multiple results for UL, mean, and LL for two time periods: 2009-2014 and 2015-2020. "mosaic.7z" contains SOC density results (Mg/ha); "mosaic_pg.7z" contains SOC stocks results (Pg), categorized by different depths (0 ~ 30 cm, 30 ~ 60 cm, 60 ~ 100 cm).

Taking 0030_2014_LL.tif as an example: 0030 represents the soil depth range (0-30 cm; other depths include 30-60 cm and 60-100 cm). 2014 refers to the time period (2014 = 2009-2014; 2020 = 2015-2020). LL represents the lower prediction limit (0.05 quantile), while UL and mean represent the upper prediction limit (0.95 quantile) and the mean estimate, respectively. The mean is used as the primary result.