ABSTRACT:

Concentration and discharge data for a high-frequency analysis of storm event dynamics at a Wetland Inlet, a Wetland Outlet, and a downstream Main Stem site in Augusta Creek, Michigan. We collected this high-frequency data beginning 25 April 2023 at the Wetland Inlet and Outlet, and 24 May 2023 at the Main Stem, with data collection ending on 28 November 2023. Concentration data includes dissolved organic carbon (DOC) and nitrate (NO3-), collected every 15 minutes using in-situ ultraviolet spectrophotometers (s::can Messtechnik GMbH, Vienna, Austria), that measure light absorbance wavelengths of 220 to 750 nm. These spectrophotometers were deployed in PVC tubes mounted in the stream channel, and powered by battery and solar panels mounted on the banks. To determine concentration values, we created site specific calibrations using partial least square regression (PLSR) models based on grab samples of DOC and NO3- collected every two weeks. Grab samples were vacuum filtered within 24 hours of field collection using 0.45 um cellulose acetate filters (Sartorius, 11106--47------N) and polysulfone filter towers triple rinsed with sample water. To measure DOC (as NPOC, non-purgeable organic carbon, mg/L), we used a total organic carbon analyzer (Shimadzu TOC-L CPH/CPN and autosampler ASI-L). To measure NO3- (as NO3--N, converted to NO3-), we used an ion chromotography system (Thermo Fisher Dionex, ICS1100 and ICS1000 and AS-DV autosampler). Our method detection limits were: 0.02 mg L-1 for nitrate and 0.43 mg L-1 for DOC. Values that are "BDL" were below instrument detection limit. For values below method or instrument detection limit, we assigned concentrations of half the limit of quantification for the purpose of calculations and analyses. If this data is used in the future, values below the method detection limit will need to be filtered appropriately. Values that are NA were not analyzed. Discharge data was measured by converting high-frequency water level data to continuous discharge data using discrete discharge measurements and a discharge-water level rating curve. Water level was calculated using pressure transducers (PT) co-located with barometric pressure loggers (Onset HOBO, Bourne Massachusetts) to determine above-sensor pressure every 15 minutes.

This data supports the paper:
Evaluating Wetland Influence on Downstream Carbon and Nitrate Dynamics During Storms in a Low-Relief Catchment
Authors: Caroline R. Weidner, Jay P. Zarnetske, Arsh Grewal, Arial J. Shogren

ABSTRACT:

Hydrology and chemistry data for a spatially distributed synoptic and discharge campaign in the Augusta Creek Catchment (located in southwestern Michigan). The spatially distributed chemistry data was collected every 2-3 weeks from October 2021 to June 2024, and includes dissolved organic carbon (DOC), nitrate, sulfate, and chloride. Within 24 hours of collection samples were filtered through 0.45µm cellulose acetate membrane filters. We measured DOC (as non-purgeable organic carbon - NPOC), using two total organic carbon analyzers with total nitrogen units (Shimadzu TOC-V CPH with a TNM-1 and autosampler ASI-V before March 9 2023 and Shimadzu TOC-L CPH/CPN with a TNM-L ROHS and autosampler ASI-L after March 9 2023). For our analysis, we only include DOC/NPOC data after March 9 2023, due to discrepancies between the two instruments used. We measured anions (Cl-, NO3-, SO42-) using an ion chromatography system (Thermo Fisher Dionex, ICS1100 and ICS1000 and AS-DV autosampler). Our method detection limits were: 0.02 mg L-1 for anions and 0.43 mg L-1 for NPOC. Values that are "BDL" were below instrument detection limit. For values below method or instrument detection limit, we assigned concentrations of half the limit of quantification for the purpose of calculations and analyses. If this data is used in the future, values below the method detection limit will need to be filtered appropriately. Values that are NA were not analyzed. We collected discharge data during a period of baseflow from August 13-14, 2024 using two Sontek Flowtracker2 Acoustic Doppler Velocimeters. Discharge values were calculated using the Flowtracker’s mid-section method, measuring depth and velocity at ~20 stations across the stream width. Uncertainty in discharge (%) was also calculated by the Flowtracker instrument. To calculate specific discharge , we normalized our measured discharge by contributing area for each flow location.

This data supports the paper:
Weidner, C. R., Zarnetske, J. P., Kendall, A. D., Martin, S. L., Nesheim, S., & Shogren, A. J. (2025). Wetlands, groundwater and seasonality influence the spatial distribution of stream chemistry in a low‐relief catchment. Journal of Geophysical Research: Biogeosciences, 130, e2025JG008989. https://doi.org/10.1029/2025JG008989

Abstract: Evaluating stream water chemistry patterns provides insight into catchment ecosystem and hydrologic processes. Spatially distributed patterns and controls of stream solutes are well‐established for high‐relief catchments where solute flow paths align with surface topography. However, the controls on solute patterns are poorly constrained for low‐relief catchments where hydrogeologic heterogeneities and river corridor features, like wetlands, may influence water and solute transport. Here, we provide a data set of solute patterns from 58 synoptic surveys across 28 sites and over 32 months in a low‐relief wetland‐rich catchment to determine the major surface and subsurface controls along with wetland influence across the catchment. In this low‐relief catchment, the expected wetland storage, processing, and transport of solutes is only apparent in solute patterns of the smallest subcatchments. Meanwhile, downstream seasonal and wetland influence on observed chemistry can be masked by large groundwater contributions to the main stream channel. These findings highlight the importance of incorporating variable groundwater contributions into catchment‐scale studies for low‐relief catchments, and that understanding the overall influence of wetlands on stream chemistry requires sampling across various spatial and temporal scales. Therefore, in low‐relief wetland‐rich catchments, given the mosaic of above and below ground controls on stream solutes, modeling efforts may need to include both surface and subsurface hydrological data and processes.