Algorithms for model parameter estimation and state estimation applied to a state-space model for one-dimensional vertical infiltration incorporating snowmelt rate as a system input
Dates
Release Date
2022-01-01
Start Date
2013-01-01
End Date
2018-12-31
Publication Date
2023-09-15
Citation
Shapiro, A.M., 2022, Algorithms for model parameter estimation and state estimation applied to a state-space model for one-dimensional vertical infiltration incorporating snowmelt rate as a system input: U.S. Geological Survey data release, https://doi.org/10.5066/P9MRGR88.
Summary
The algorithms and input data included in this data release are used to interpret time-series data (water-table altitude, precipitation, snowmelt, and potential evapotranspiration) over an observation period to estimate model parameters of a State-Space Model (SSM) of vertical infiltration to the groundwater table. The SSM model is coupled with a Kalman Filter (KF) to estimate system states (water-table altitude and groundwater recharge) over the observation period. This SSM and KF model is formulated for one-dimensional vertical infiltration and includes preferential and diffuse flow through the unsaturated zone to the water table. The analysis was conducted to demonstrate the application of the SSM and KF model in characterizing [...]
Summary
The algorithms and input data included in this data release are used to interpret time-series data (water-table altitude, precipitation, snowmelt, and potential evapotranspiration) over an observation period to estimate model parameters of a State-Space Model (SSM) of vertical infiltration to the groundwater table. The SSM model is coupled with a Kalman Filter (KF) to estimate system states (water-table altitude and groundwater recharge) over the observation period. This SSM and KF model is formulated for one-dimensional vertical infiltration and includes preferential and diffuse flow through the unsaturated zone to the water table. The analysis was conducted to demonstrate the application of the SSM and KF model in characterizing responses of the groundwater table and estimating time-varying groundwater recharge due to the combination of liquid precipitation events and snowmelt events. The SSM and KF are applied to daily information for water-table altitude, liquid precipitation, snowmelt, and potential evapotranspiration. In fractured rock aquifers, rapid infiltration to the groundwater table following precipitation and snowmelt events may result in groundwater contamination from surface contaminants or pathogens. The magnitude of the time-varying groundwater recharge can be used as a surrogate to indicate time-varying contamination susceptibility of the groundwater, as microbial, particulate, and other groundwater quality chemical indicators are unlikely to be available or are costly to develop with the temporal frequency needed to resolve responses to precipitation and snowmelt events. The SSM and KF can capitalize on currently available technologies and telecommunication infrastructure that deliver real-time data for water-table altitudes and meteorological inputs to conduct real-time recharge estimation. Simulations are conducted to demonstrate the application of the SSM and KF to the interpretation of time-series data for water-table altitude, liquid precipitation, snowmelt, and potential evapotranspiration. The data used in this demonstration are from a period of record in between January 1, 2013 and December 31, 2018 for groundwater monitoring wells in northwestern New York, USA; the groundwater monitoring wells measure groundwater responses in the carbonate aquifers that are used for watersupply in this region. Water-table altitude data are available from the U.S. Geological Survey (USGS), and meteorological data are available from the National Oceanic and Atmospheric Administration (NOAA), and the Snow Data Assimilation System (SNODAS). Seasonal (Spring, Summer, Fall, and Winter) simulations using the SSM and KF are presented in this data release. The algorithms used to formulate the SSM and KF and interpret the time-series data are prepared in the software MATLAB, where functional calls are made to available algorithms that conduct the parameter estimation of the SSM parameters, followed by the application of the KF to perform the estimation of model states. The MATLAB files developed for the simulations are available in this data release. MATLAB is a proprietary software, and thus, a stand-alone and executable version of the algorithms is not available in this data release. Details regarding the availability of MATLAB are available from https://www.mathworks.com. This USGS data release contains all input and output files for the simulations described in the associated journal article (https://doi.org/10.1111/gwat.13206).
The algorithms and data presented in this data release were developed to demonstrate the application of a State-Space Model (SSM) of one-dimensional vertical infiltration that interprets time-series data for water-table altitudes, liquid precipitation, snowmelt, and potential evapotranspiration to estimate time-varying groundwater recharge. The SSM is coupled with the Kalman Filter (KF) to perform the estimation of the system states. The SSM of infiltration incorporates preferential and diffuse flow to the groundwater table through the unsaturated zone and estimates model parameters for an observation period. Observation periods are defined as seasons (winter, spring, summer, and fall), where the model parameters are assumed to be constant of the observation period. Model parameters can vary from one observation period to the next. The SSM and KF are demonstrated on groundwater monitoring wells in northwestern New York, USA, where both liquid precipitation and snowmelt in the fall, winter, and spring seasons contribute to groundwater recharge. The development of the model input and output files included in this data release are documented in the journal article available at https://doi.org/10.1111/gwat.13206.
Preview Image
Image of the map location where estimates of groundwater recharge are conducted.