Lake Kegonsa Agrivoltaic Dataset

Ancillary micrometeorological, solar-radiation-component, within-array PPFD light-bar, and soil moisture / temperature observations supporting the paired Lake Kegonsa agrivoltaic flux towers (US-KSA and US-KSC), Wisconsin, USA, spanning the 2025 growing season.

Note on the eddy-covariance / flux tower data: the half-hourly EC + micromet records from the US-KSA (agrivoltaic) and US-KSC (open control) towers are not stored in this HydroShare resource. They are distributed via the AmeriFlux repository; see fluxTower/README.md for the canonical URLs and download instructions. The fluxTower/ directory in this bundle is intentionally a documentation-only placeholder.

Sites and platforms

• US-KSA Agrivoltaic flux tower (hosted on AmeriFlux; see fluxTower/)
• Eddy-covariance tower under a single-axis tracker PV array.
• US-KSC Open-field control flux tower (hosted on AmeriFlux)
• Shadow-band radiometer (CR1000 logger): hourly direct / diffuse / total solar radiation partitioning. Co-location with US-KSA / US-KSC towers: TBD.
• Light-bar PPFD network: eight LI-COR quantum line sensors deployed across both panel systems (Single-axis tracking and Fixed-tilt) at Under-panel, Row-edge, and Alley positions; 5-min sampling.
• Soil moisture / temperature network: eight METER RXW-GP4A profile probes distributed across alley, below-panel, dripline, and control positions for both panel systems; volumetric water content at 4 depth bands (0-15, 16-30, 31-45, 46-60 cm) and point temperatures at 6 depths (3.5, 10, 20, 30, 45, 60 cm). 10-min sampling.

File organization (manifest)

Top-level directories group sources by platform; within each, data are subdivided by measurement family.

fluxTower/ empty placeholder - see README inside README.md redirect to AmeriFlux US-KSA / US-KSC shadowband/ shadow-band radiometer, hourly, 2025-06-02 to 2025-10-10 shadowband_radiation.csv Shadow_Avg, Clear_Avg, Bar_Avg, Shadowfiveminavg_Avg, Stot, Sbeam, Sdiff lightbar_ppfd/ LI-COR light bars, 5 min, 2025-05-30 to 2025-10-10 sensor_inventory.csv per-sensor metadata (8 rows) lightbar_ppfd_tracking.csv L1-L4 (single-axis tracking system) lightbar_ppfd_fixed.csv L5-L8 (fixed-tilt system) soil/ METER RXW-GP4A probes, 10 min soil_sensor_inventory.csv per-probe metadata (8 probes) soil_moisture_10min.csv 32 SM channels, May 2025 to Jan 2026 (Dec gap) soil_temperature_10min.csv 48 T channels, May to Nov 2025 soil_moisture_correction.md documentation of temperature correction used for the manuscript-figure variant

Each data CSV begins with a commented metadata header (lines starting with #) followed by a single column row and the data records. To read in pandas:

python import pandas as pd df = pd.read_csv('soil/soil_moisture_10min.csv', comment='#', parse_dates=['datetime_iso'])

Companion files

• variables.csv - controlled vocabulary mapping each short variable code to a long name, CF/CUAHSI standard name, units, instrument, and the sites it applies to.
• methods.md - flux processing chain, QC scheme, gap-filling notes, and outstanding items (TBD fields).
• CHANGELOG.md - version history.
• LICENSE - CC-BY-4.0 license text.

Time convention

All timestamps are ISO-8601 with an explicit offset.

• shadowband: -06:00 (America/Chicago Local Standard Time, no DST applied). Each timestamp marks the start of the 1-h averaging interval.
• lightbar_ppfd: Z (UTC). Timestamps mark the sample instant. The source files use the UTC convention; preserved as-is here to avoid lossy round-trips. Convert downstream if needed: UTC - 6 h = LST.
• soil: the source export embeds an explicit offset per timestamp (-05:00 for CDT, -06:00 for CST). The offsets are preserved verbatim - note that the November 2025 source file remained on -05:00 past the 2025-11-02 DST transition while the January 2026 source file moved to -06:00. Normalise downstream if a single canonical offset is required.

Missing values

Missing values are coded as NA. No gap-filling has been applied in this release; downstream users should apply their preferred gap-filling / partitioning workflow (e.g., REddyProc, ONEFlux).

Record counts

fluxTower (US-KSA, US-KSC)

• External - hosted on AmeriFlux. See fluxTower/README.md for retrieval. The directory in this bundle is an empty placeholder.

shadowband (hourly)

• shadowband_radiation.csv: 3132 records

lightbar_ppfd (5 min)

• sensor_inventory.csv: 8 sensors
• lightbar_ppfd_tracking.csv: 66857 records (L1-L4)
• lightbar_ppfd_fixed.csv: 90739 records (L5-L8)
• Note: 128 of 157596 records are off the nominal 5-min grid (mostly off-by-one-second; plus a small instrument-startup cluster on 2025-05-30). Preserved as-is - filter downstream if a strictly regular grid is required.

soil (10 min)

• soil_sensor_inventory.csv: 8 probes
• soil_moisture_10min.csv: 31691 records x 32 SM channels (8 probes x 4 depth bands), 2025-05-20 to 2026-01-31
• soil_temperature_10min.csv: 27798 records x 48 T channels (8 probes x 6 point depths), 2025-05-22 to 2025-12-01
• Note: December 2025 is absent from the SM source export (gap between the November and January files); soil temperature was not exported for December 2025 or January 2026. Values of 0.0 in the early record (May-Jun 2025) typically indicate pre-deployment / pre-calibration. No filter applied.

Citation

Han, K., Mackin, H., Mather, E., Lawton, S., Thom, J., Desai, A. R., Kucharik, C. J., Loheide, S. P. (2026). Lake Kegonsa agrivoltaic ancillary dataset - shadow-band radiation, within-array PPFD, and soil moisture / temperature (2025 growing season). HydroShare. DOI: TBD

For the paired eddy-covariance flux records (US-KSA, US-KSC), cite the AmeriFlux site data products separately following the AmeriFlux Data Use Policy.

Contact

khan99@wisc.edu

License

Released under Creative Commons Attribution 4.0 (CC-BY-4.0).