Incorporating Snowmelt into Daily Estimates of Recharge Using a State-Space Model of Infiltration

Area of Science:

• Hydrology
• Hydrogeology
• Water Resource Management

Background:

• Groundwater recharge estimation is crucial for water resource management.
• Traditional methods often lack the temporal resolution and data integration capabilities for real-time assessments.
• Understanding infiltration dynamics, including diffuse and preferential flow, is key to accurate recharge modeling.

Purpose of the Study:

• To develop and demonstrate a state-space model (SSM) coupled with a Kalman Filter (KF) for daily groundwater recharge estimation.
• To assimilate time-series of groundwater-level altitude and meteorological data for improved recharge predictions.
• To identify potential discrepancies between meteorological inputs and observed hydraulic head data.

Main Methods:

• A state-space model (SSM) was developed to simulate infiltration, incorporating diffuse and preferential flow pathways.
• The SSM was integrated with the Kalman Filter (KF) to assimilate real-time hydraulic head and meteorological observations.
• Model parameters were estimated seasonally, and the model was applied to daily data from two bedrock wells in New York.

Main Results:

• The model successfully estimated daily groundwater recharge, capturing preferential flow events during winter and spring from precipitation and snowmelt.
• Annual recharge estimates aligned with those from previous groundwater flow and surface-process models.
• Discrepancies between modeled snowmelt rates and observed hydraulic head indicated potential errors in meteorological input data.

Conclusions:

• The coupled SSM and KF approach provides a robust method for real-time groundwater recharge estimation.
• The study highlights the importance of accurate meteorological data, particularly snowmelt, for reliable recharge modeling.
• The model's ability to detect input data inconsistencies offers a valuable tool for data quality assessment in hydrological studies.