A New Model for Synthesis: The Soil Warming Experiment to Depth Data Integration Effort (SWEDDIE)

Ecosystem warming experiments offer key insights into the functioning of plants, microbes, and biogeochemical processes in a warmer world, but are limited by their ecological context, within site heterogeneity, and instrumentation. The Soil Warming Experiment to Depth Data Integration Effort (SWEDDIE) is a joint platform developed by the DeepSoil2100 network (23 warming experiments worldwide) to overcome these limitations through the creation of a network-wide database.

The SWEDDIE database is designed with FAIR principles to accommodate a wide range of data types, with a streamlined data ingestion system and user-friendly query and reporting tools. Comprehensive metadata reporting standards enable SWEDDIE to serve as a repository for past, present, and future datasets. Harmonization of data is facilitated with data dictionary files that accompany and describe variables in each data file and also store sensor and methods information. SWEDDIE consists of both publicly accessible and network-only data tiers, and is hosted on ESS-DIVE to leverage existing data repository infrastructure. Data ingestion, wrangling, and synthesis tools for SWEDDIE are also available in a companion R package.

A key tenet of SWEDDIE is the inclusion of soil measurements below 0.2 m, as the warming response of C stocks in deeper soil layers remains both highly uncertain and consequential for the global C cycle. Accordingly, we welcome new sites and data submissions, provided that > 1 °C warming has been observed below this depth. The first synthesis analysis with SWEDDIE focused on the impact of warming on soil moisture, while also serving to test the SWEDDIE data model and refine the harmonization approach. The results of this analysis provide quantitative evidence that warming leads to decreased soil moisture throughout the soil profile, but with more drying in surficial compared to deeper soil layers. The degree of warming correlates directly with the magnitude of soil drying, but the specific relationship between warming and drying varies by site as well as seasonally.  

SWEDDIE is at its core a community project. The results of the preliminary soil moisture analysis are a key building block of one of the next planned synthesis efforts: using selected experiments to benchmark soil C warming responses with the ecosys model. This demonstrates the positive feedback inherent to this platform, i.e., that active community engagement leads to improved data coverage, which in turn enhances our capacity to generalize warming responses across ecological gradients, inform global models, and quantify potential experimental biases.