Scholarlink

<p>Quantifying groundwater recharge at a daily temporal resolution is difficult when inputs are sparse. This research presents a simple lumped model that links precipitation to groundwater levels by (1) routing rainfall through a first-order exponential vadose filter followed by an explicit discrete delay, and (2) representing lateral drainage with a nonlinear (quadratic) outflow law. The saturated storage is tracked with specific yield (<i>S</i><sub>y</sub>), and the model advances at a daily time step, producing time series of recharge, head, and lateral flow. The approach was tested with daily precipitation data from nearby rainfall stations and groundwater levels from four wells in the Bauru Aquifer System (BAS, Brazil). The model reproduces the timing and amplitude of observed hydraulic heads across seasonal cycles, capturing delayed and attenuated recharge pulses and curvilinear recessions, with a normalized root mean square error (RMSE) &lt; 14% against observations. Relative to the water-table fluctuation (WTF) method, the mass balance explicitly accounts for concurrent lateral drainage during recharge events and for vadose delay/attenuation, explaining systematically higher, but process-consistent, recharge fractions than the WTF method. Parameter estimates are hydrologically interpretable, although equifinality requires validating fitted values against independent aquifer information. The method provides fast recharge diagnostics from long daily records without requiring daily evapotranspiration, making it suitable for basin-scale screening and trend attribution, and extensible to multistress forcing where data permit.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A lumped model with vadose filtering and quadratic drainage for daily groundwater recharge estimation

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月20日

查看全文
  • Caio Marques Manara,
  • Elias Hideo Teramoto,
  • Gustavo Soldado Peres,
  • Pedro Luiz Becaro Crioni,
  • Hung Kiang Chang

摘要

Quantifying groundwater recharge at a daily temporal resolution is difficult when inputs are sparse. This research presents a simple lumped model that links precipitation to groundwater levels by (1) routing rainfall through a first-order exponential vadose filter followed by an explicit discrete delay, and (2) representing lateral drainage with a nonlinear (quadratic) outflow law. The saturated storage is tracked with specific yield (Sy), and the model advances at a daily time step, producing time series of recharge, head, and lateral flow. The approach was tested with daily precipitation data from nearby rainfall stations and groundwater levels from four wells in the Bauru Aquifer System (BAS, Brazil). The model reproduces the timing and amplitude of observed hydraulic heads across seasonal cycles, capturing delayed and attenuated recharge pulses and curvilinear recessions, with a normalized root mean square error (RMSE) < 14% against observations. Relative to the water-table fluctuation (WTF) method, the mass balance explicitly accounts for concurrent lateral drainage during recharge events and for vadose delay/attenuation, explaining systematically higher, but process-consistent, recharge fractions than the WTF method. Parameter estimates are hydrologically interpretable, although equifinality requires validating fitted values against independent aquifer information. The method provides fast recharge diagnostics from long daily records without requiring daily evapotranspiration, making it suitable for basin-scale screening and trend attribution, and extensible to multistress forcing where data permit.