We defined water flow regulation (WFR) as the ability of forested catchments to maintain streamflow within an acceptable range defined by the needs of the beneficiary populations. This service has two main components, drought prevention, or the maintenance of water above a low flow threshold, and flood prevention, the maintenance of water below a high flow threshold. At HBEF, the high flow threshold was chosen based on the structural capacity of the Franklin Falls Dam (Figure 1). This dam, maintained by the U.S. Army Corps of Engineers, controls the outflow of the Pemigawasset River watershed into which Hubbard Brook ultimately drains. Based on technical documentation, we determined the maximum daily discharge of the dam and translated this volume of water to a height over the entire area of the Pemigawasset River watershed (Figure 2). This value constitutes the average maximum daily area-weighted streamflow that catchments within the watershed can discharge before the dam exceeds its maximum discharge. This value was used as the high flow threshold. The low flow threshold was determined in a similar fashion, based on historical water use data provided by the U.S. Geological Survey (USGS). In the case of HBEF, the annual volume of surface water extracted from Grafton County was translated to a daily height over the entire area of the county. This value constitutes the average minimum area-weighted streamflow that catchments must maintain before daily surface water use exceeds supply. This value was adopted as the low flow threshold. USGS water-use records are reported on a 5-year time step starting in 1985, and we assumed water use in a given year to be equivalent to the value of water use for the most recent record prior to that year. For all dates prior to 1990, we used the 1985 value. As measures of the service, we calculated the percentage of days each year in which streamflow remained within the flow thresholds, as well as the mean duration of the period between subsequent threshold events. At other sites, we used values relevant to local demand. Where data describing demand was not available, we set high and low thresholds to values equal to statistical quantiles (e.g. 5th and 95th for high and low thresholds, respectively) based on the total dataset for flows at that site. Once thresholds were set, we calculated the percentage of measurements days in which flow remained below the high threshold (flood regulation), above the low threshold (drought mitigation), and between both thresholds (the "Goldilocks Zone" or GLZ, Figure 3), as well as the duration of time between subsequent threshold events (flow stability).
FEST case studies involving this service include quantification of WFR at Hubbard Brook and Turkey Lakes, understanding the impact of different thresholds and different beneficiary groups at Frost Valley, Hubbard Brook, and Turkey Lakes, and exploring the role of climate forcing on service provision, as well as the inclusion of WFR metrics in tradeoff analysis.