Profiling equipment: Dissolved oxygen (optical, luminescence; accuracy: ±0.1 mg/L, resolution: 0.01 mg/L) and temperature (accuracy: ±0.01°C, resolution: 0.001°C) profiles were continuous and collected using a Yellow Springs Instrument (YSI) EXO3. In addition, the EXO3 included sensors for: depth (stainless steel strain gauge, resolution: 0.001 m), conductivity (4-electrode nickel cell, resolution: 0.1 microsiemens/cm to 10 microsiemens/cm, range dependent), pH (glass combination electrode, resolution: 0.01), oxidative-reductive potential (ORP, platinum button, resolution: 0.1 mV), chlorophyll a (optical, fluorescence, resolution: 0.01 RFU), phycocyanin (PC, optical, fluorescence, resolution: 0.01 RFU), phycoerythrin (PE, optical fluorescence, resolution: 0.01 RFU), and turbidity (optical, 90° scatter, resolution: 0.01 FNU).
Data quality assurance/quality control (QA/QC): From collection, the data for each profile was carefully QA/QC’ed using automated processes to ensure data quality. First, if the profile had more than 400 readings, the sonde was lowered very slowly with likely replicate readings at different depths. The data, therefore, was averaged across all parameters by 3 second bins. Next, all rows with depths recorded as ≥ 0m were deleted as that would indicate readings from above the water surface. Any profile readings were deleted with a depth difference from one sequential depth to the next of ≤ 0.02 m. This step eliminated locations in the profile where the sensor was pulled upward, was stationary, or had settled at the bottom of the waterbody. All remaining data points were checked for error codes or data outside the sensor bounds as specific by the instrumentation and were converted to NA. Next, all water temperature values were converted to density and densities profiles were checked for any non-monotonically decreasing densities. All temperature values were checked for values that were within the sensor limits but were unrealistic (e.g., < 4.2°C outside of November to April). Those profiles were removed in their entirety. All turbidity, chlorophyll a, phycocyanin, and phycoerythrin fluorescence were checked for large jumps (3x) in the bottom 15% of rows. If there was a large jump in any of those values, the sensor had entered the transition layer at the sediment/water interface consisting of loosely packed colloidal suspensions (“fluffy layer”, Strakhovenko et al. 2020). The rows of data were deleted from the first jump in the bottom 15% of data down to the lowest point. The maximum depth for these profiles is considered the new maximum depth after the preceding QA/QC steps. Any profiles were excluded if there were less than 5 readings in the profile. Each year has a single data file for all profiles taken in that year. We provide metadata (latitude and longitude) of the sampling locations for all reservoirs sampled for this dataset and other reservoirs sampled regularly See Jones et al. (2023) for previous years profile data.
Citations:
Jones, JR, RL North, A Argerich, AP Thorpe, DV Obrecht, A Price, K Santamaria, and DC Richardson. 2023. Missouri reservoir profile data including temperature, depth, and oxygen profiles (1989-current) ver 1. Environmental Data Initiative. https://doi.org/10.6073/pasta/34d23adf277b07346a35fb8cdc984cc7 (Accessed 2023-12-18).
Strakhovenko, V., Subetto, D., Ovdina, E., Danilenko, I., Belkina, N., Efremenko, N., & Maslov, A. 2020. Mineralogical and geochemical composition of Late Holocene bottom sediments of Lake Onego. Journal of Great Lakes Research, 46(3), 443-455.
