Soil water content and water salinity (DP1.00094.001) algorithm updates

Soil water content and water salinity (DP1.00094.001) is being migrated to NEON's new instrument processing pipeline. As part of this migration, several longstanding problems with the data product will be addressed and provisional data will be generated using 5-minute and 30-minute averaging intervals, replacing the 1-minute and 30-minute averaging intervals used in RELEASE-2026 and earlier. 

The change from 1-minute to 5-minute averaging intervals is based on feedback from NEON's Soil Sensor Technical Working Group and other external researchers who found the 1-minute data volume too large to be useful and noted that soil moisture does not typically change appreciably from one minute to the next. This change will create a temporary misalignment between newly generated provisional data (5-minute/30-minute) and previously released data (1-minute/30-minute). Data users who combine provisional and released Soil water content and water salinity data during this period should be aware of this discrepancy in the averaging interval. 

All Soil water content and water salinity data — both provisional and previously released — will be reprocessed to the 5-minute and 30-minute averaging interval and made available in RELEASE-2027, expected in late January 2027. After RELEASE-2027, the averaging interval will be consistent across the full period of record.

Prior to the publication of RELEASE-2027 data, users downloading both released (includes 1-minute data) and provisional (includes 5-minute data) data will need to use neonUtilities R package version 4.1.0 or later, which is expected to be released in November 2026. Until version 4.1.0 is available, users can avoid this issue by restricting downloads to the 30-minute data (i.e., timeIndex = 30 in the loadByProduct function) or by accessing released and provisional data separately.

Other changes with the new processing pipeline include: 

• Sensor measurement depth will be correctly reported in the sensor positions file (negating the need for the workaround swc_depthsV3 file on the data product webpage)
• Soil moisture will be reported based on both the factory calibration and a soil-specific calibration (except permafrost sites)
  • For users needing relative soil water content data (e.g., identifying drier and wetter periods), both the factory calibration and soil-specific calibration are expected to perform similarly. For users needing absolute soil water content data, the soil-specific calibration has a lower measurement uncertainty than the factory calibration and, in most cases, should be more accurate. Instances where the soil-specific calibration may not be as accurate as the reported uncertainty could occur if the location where the sensor was installed was not well represented by the nearby location where soil was collected to determine the soil-specific calibrations, but this is expected to be rare. Additionally, soil-specific soil water content calibrations are not available at permafrost sites (HEAL, BONA, BARR, and TOOL) as it was not possible to collect the large intact blocks of soil needed to generate them; instead, only soil water content based on the factory calibration is available at these sites.
• The temperature test will not be implemented at times and places where the soil is very unlikely to be frozen
• The automated QA/QC thresholds have been adjusted

A sizable increase in the availability of unflagged soil water content data is anticipated due to changes to the automated QA/QC flagging thresholds and the restriction of the temperature test to times and places where the soil could plausibly be frozen. Most suspect data will still be flagged appropriately.

All code used to generate the Soil water content and water salinity data product is open-source and available in the NEONScience/NEON-IS-data-processing GitHub repository. 

For more information on this update, please use the NEON Contact Us form and reference DP1.00094.001.