Long-Term Radon-222 (222Rn) and Hydroclimatic Dataset for a Coastal Estuary, Corpus Christi Bay, Texas

Published: 29 August 2023| Version 1 | DOI: 10.17632/s9m8t7fg4k.1
William Wolfe


Radon-222 (222Rn) is a naturally occurring radio-active tracer commonly used to estimate submarine groundwater discharge (SGD). "radon raw dataset.csv" contains 222Rn in water collected periodically from Aug. 2019 to June 2021 on 30-minute intervals at a near-shore monitoring platform in Corpus Christi Bay, TX, USA (n= 10,660). Also included in this file are continuous, publicly available hydroclimatic parameters (wind speed and direction, atmospheric pressure, water temperature, tide height, creek and river discharge rate, n=35,088). The raw data set was input into two machine learning model scripts: Random forest (Radon_VarTransform_RFprediction_GAM.R) and Deep Neural Network (DNN) (Radon_DNNprediction.txt). The models explain radon variability and predict over gap periods. A generalized additive model (GAM) was also employed to explain 222Rn inventory variability modulated specifically by windspeed and direction. The data and modelling tools included in this article provide new opportunities to 1) predict radon inventories in regions with large temporal data gaps using only publicly available hydroclimatic parameters and 2) better constrain SGD estimates in windy coastal areas, allowing for insights that can be used to plan field excursions and management strategies for coastal water and solute budgets.


Steps to reproduce

Steps are defined in the .R file and the Data in Brief article associated with Wolfe et al. 2013 https://doi.org/10.1016/j.jhydrol.2023.130065


Groundwater, Machine Learning, Radon, Tracer Isotope