Abstract:This study is aimed at quantifying the difference in aquifer's response to recharge between some different locations in a fan aquifer and a delta aquifer for a preliminary study of revealing mechanisms of water transport in alluvial aquifer. The aquifer's response to recharge is statistically quantified with the two viewpoints: (1) timing and volume of recharge and (2) time length of aquifer's holding water. For the first point, a statistical model that links precipitation and groundwater level is introduced, and its parameters are identified using correlation analysis. Our results show that the recharge rate at the toe is higher than that at the apex and at the delta. For the second point, the concept of 'memory effect' of aquifer is adopted and quantified using the autocorrelation and spectral analyses. Our results show that the memory effect is longer at the toe of fan than at the apex, and thus, a temporary increase of water level has about five times as long-term influence on subsequent water levels at the toe of the fan as at the apex. This study demonstrates that the statistical analyses and modeling of hydrological data are useful for characterizing aquifer's hydrodynamics.

A distributed hydro-environmental model is developed that achieves detailed analysis of the movement of water at a field-plot-scale resolution in a mesoscale watershed including lowland areas where, especially for agricultures, it is an essential need to get rid of redundant groundwater by drainage facilities such as rivers, canals and/or underdrains. For this, the problem geometry is meshed with unstructured cells of triangular shape. Profile of a column cell is zoned into two: surface zone and groundwater zone in which water movement is represented by combined tank and soil moisture sub-models, and welldefined two-dimensional unconfined shallow groundwater flow sub-model, respectively. The top-two sub-models serve to evaluate evapotranspiration, infiltration, soil water content, lateral surface water flow, and vertical percolation. The vertical percolation so evaluated is given as longitudinal recharge to the bottom sub-model for computing groundwater flow. Surface water-groundwater interactions through beds and stream-banks of perennial and ephemeral canals are considered by treating the canal courses as internal boundaries in the groundwater flow model. The finite volume method (FVM) that allows of unstructured mesh and produces conservative solutions is employed for groundwater flow computation. The model developed is applied to an actual watershed which includes a low-lying paddy area to quantify the hydrological impact of land-use management practices over a period of 29 years in which the farmland consolidation project was implemented and part of the paddy fields were converted to upland crop fields and housing lands. From the results obtained, it is concluded that the model presently developed lends itself to water-as well as land-use management practices.

An optimization model for cropping-plan placement on field plots is presented for supporting decision-making on agricultural management by a farming organization. The mixed 0-1 programming technique is employed to select the next planting crop at each field plot in a holistic manner. Reduction of total nitrogen discharged from field plots to the downstream end of the drainage canals is expressed as an objective function of the model to balance an achievement of economic goal and environmental conservation. Some Japanese governmental policies on regulating rice cropping areas and on promoting production of particular upland field crops can be formulated in the model. A computational example of cropping-plan placement on field plots managed under integrated policies is given by operating the optimization model with various weights associated with the objectives. The procured tradeoff curve and corresponding patterns of cropping-plan could be useful in the decision-making by the farming organization.

A sophisticated modeling approach for simulating-coupled surface and subsurface flows in a watershed is presented. The watershed model developed is a spatially distributed physically based model of composite dimension, consisting of 3-D variably saturated groundwater flow submodel, 2-D overland flow submodel and 1-D river flow submodel. The 3-D subsurface flow is represented by the complete Richards equation, while the 2-D and 1-D surface flows by the diffusive approximations of their complete dynamic equations. For piecewise integration of these equations, the finite volume method (FVM) is employed assuming unknown variables such as the water depth and the pressure head to be volume-averaged state ones. Problem plane geometry is meshed with the unstructured cells of triangular shape which conforms to external as well as internal irregular boundaries such as those between 1-D and 2-D flows. A cell size controlling scheme, referred to as quasi-adaptive meshing scheme, is introduced to keep the local discretization errors caused by topographic elevation gradient even over the entire-meshed geometry. Performance of the model is tested through its practical application to a rugged intermountain watershed. Tuning the values of the three key parameters ensures successful calibration of the model. Once the model is so calibrated, it could reproduce satisfactory runoff response to any rainfall event. Expansion and shrinkage of the contributing area importantly affecting the direct runoff, caused by the vicissitude of rainfall during its total duration, are well reproduced, like what the commonly accepted runoff theory argues. It is thus concluded that the model developed could serve as a powerful watershed simulator usable for investigating and assessing the hydrological aspect of a watershed.

Long-term simulation using the distributed hydro-environmental watershed model is efficacious for assessing irrigation impacts on hydrological cycle in detail and for implementing watershed management successfully. In this article, the previously developed hydro-environmental watershed model (HEWM-1) is improved in the water exchange process caused by surface water-groundwater interaction via drainage canals and/or underdrains. The time-varying stream flow in canals is described by the complete one-dimensional shallow water equations in a newly introduced submodel, the open channel flow submodel. This submodel coordinates with the other submodels: the tank, soil moisture and groundwater flow submodels which are interlinked in a cascade manner. The improved model (HEWM-2) is applied to an agricultural watershed covering an area from an alluvial fan onto a nearly level alluvial plain, to be validated. The simulation by HEWM-2 is informative for identifying whether any drainage canal is gaining or losing water in relation to groundwater level. It could thus provide useful information for conserving a complex network of drainage canals which also functions as a passage for aquatic animals like fishes.