Figure 2 . Photographs of some field locations and infrastructure installations in the ER: (A) Hillslope borehole, subsurface water and carbon inventory well (B) Weather station, atmospheric weather sensor in the alpine zone, (C) Pumphouse ISCO, sampler for subsurface discharge and solute flux in the subalpine zone, (D) A plot to study impacts of early snowmelt on vegetation at a subalpine location along an elevation gradient
The ER watershed also has infrastructure maintained by several federal, state, and local agencies that have different data systems (some of these are indicated in Figure 1a). Notably, the Rocky Mountain Biological Laboratory (RMBL, https://www.rmbl.org/) is situated in the townsite of Gothic, and has over 90 years of data collection activities in the watershed. Snow measurements and associated meteorological data are available from the National Resources Conservation Service (NRCS) snow Telemetry (SNOTEL) sites ‘Butte’ and ‘Schofield Pass’, Crested Butte Cooperative Observer Network (COOP), and a number of Weather Underground stations. The USGS maintains gaging stations (located downstream of the SFA gages), and collects water quality measurements across the East-Taylor watersheds, and makes the data available through NWIS (HUC: 14020001). Additional water quality data are available from the National Water Quality Portal, which includes measurements by the U.S. Environmental Protection Agency (EPA), Colorado Department of Public Health and Environment (CDPHE), and local groups including the Coal Creek Watershed Coalition and the Rivers of Colorado Water Watch. The EPA also maintains a National Atmospheric Deposition Program (NADP) and Clean Air Status and Trends Network (CASTNET) station at Gothic.
Thus collectively, the SFA and its collaborators generate diverse, multiscale datasets at the ER including hydrological, (bio)geochemical, climate, vegetation, geophysical, microbiological, and remote sensing data. Additionally, several model datasets, including inputs, outputs and preprocessing codes are generated from numerical simulations of different watershed subsystems and their aggregated behavior. Detailed descriptions of the data types, data variables collected, and methods used are listed in Table 1. The most common publicly available data types from the ER during 2015-2020 are biogeochemistry and hydrology data.
The SFA has a Data Management Framework component that provides services and infrastructure to support the project’s data lifecycle. The framework comprises systems, workflows and scripts to acquire and store data in a queryable database, conduct QA/QC, integrate project data with external data for real-time queries, discover and download data, and to publish data with digital object identifiers (DOIs) in the DOE’s Environmental Systems Science Data Infrastructure for a Virtual Ecosystem (ESS-DIVE) repository (Varadharajan et al., 2019). Additionally, the SFA has developed an integrated field-data workflow to acquire critical metadata from these diverse data streams, and to manage data at each stage of the scientific process. The field-data workflow outlines guidelines for managing metadata and identifiers for field locations, samples, sensors, and data package creation and publication. While developed by the SFA, the workflow is available to other collaborating organizations, and is built based on community feedback and established data management best practices.
4. Scientific Impact of ER Datasets
The multidisciplinary data from the ER have advanced the understanding of hydrological processes in mountainous catchments, and the simultaneous collection of diverse data types across the watershed allows researchers across institutions to share knowledge and draw conclusions. Examples of datasets collected through periodic sampling or from sensor instrumentation that are frequently used in analysis and modeling include daily river stage observations and discharge estimates (dataset 43; observed measurement errors ranging from ~2-15% ), weekly-daily surface and groundwater geochemistry (e.g. datasets 3, 5, 9, 15, 21, 41; measurement uncertainties in Tables S2-S3), and instantaneous groundwater elevations and temperature in the hillslope and floodplain (e.g. datasets 30, 46). Examples of research topics that ER datasets are used for include water partitioning as a function of hydrological perturbations such as drought and early snowmelt (e.g. datasets 18, 26, 39-44), surface-groundwater interactions (e.g. datasets 63-68), nitrogen cycling and export to the river (e.g. datasets 15, 20, 21, 27, 31-32, 52, 56), the impact of weathering and other processes on river water composition (e.g. datasets 10, 28, 32), and more broadly an aggregated understanding of watershed hydrobiogeochemical processes given their variation over space and different watershed functional zones and compartments (e.g. datasets 49-54).
The ER datasets have been used in numerous publications of which a select few are highlighted here. For example by combining measurements of river, rain, groundwater and snow chemistry, stream discharge, remote sensing (LIDAR, ASO), Carroll et al. (2018, 2019) found that groundwater recharge, an important contributor to streamflow, is dependent on elevation and vegetation (datasets 42-44, 51, 70-72). Specifically, groundwater recharge increases in higher elevations, such as the upper subalpine zone where there is greater snow accumulation and lower canopy cover. Through analyses of data on groundwater chemistry, water table depth, and rock mineralogy, Wan et al. (2019) found that the seasonal water table depth determines the weathering zone and weathering front in sedimentary bedrock, and that the Mancos shale can be a significant contributor to river nitrogen exports (datasets 30-32). Combining snow measurements with metagenome analysis, Sorensen et al. (2020) found that snowmelt triggers a pulse of nitrogen in hillslope soils concomitant with a collapse in microbial biomass, and changes in microbial community composition (datasets 17, 20). Using model simulations of floodplain meanders and regions of hyporheic exchange, Dwivedi et al. (2017, 2018) found these subsystems exert critical controls on nitrogen cycling and other solute exports to the river (datasets 29, 46-47).
The interdisciplinary data from the ER can also be used in future investigations that address science questions identified by the broader community such as the impacts of climate change and extreme events on the critical zone, and the scale dependence of hydrology (Blöschl et al. 2019). These data along with future measurements from the SAIL campaign will provide integrated observational datasets for benchmarking atmospheric and hydrological models in mountainous watersheds, thus addressing an identified data gap in modeling mountain rain and snow (Lundquist et al. 2019). More broadly the data from the ER will help understand the impacts of hydrological perturbations on water availability and quality in the Western United States.
The findings from the ER community, and the potential for gaining future scientific insights using the data, highlight the value of designing multidisciplinary watershed observatories using open science by design principles, and publishing data generated in open, public repositories (Stegen et al., 2019).