Soil water infiltration models testing, with the use of experimental and computational simulation for surface and groundwater management
Δοκιμή μοντέλων διήθησης νερού στο έδαφος, με χρήση πειραματικής και υπολογιστικής προσομοίωσης για διαχείριση επιφανειακών και υπογείων υδάτων

Doctoral Thesis
Author
Giakoumatos, Stefanos D.V.
Γιακουμάτος, Στέφανος
Date
2025-06Advisor
Sidiras, DimitriosΣιδηράς, Δημήτριος
View/ Open
Keywords
Διαχείριση υδάτων ; Ποιότητα υδάτων ; Νιτρορύπανση εδάφους ; Νιτρορύπανση υδάτων ; Μοντέλα καθίζησης ; Συμπίεση ιλύοςAbstract
The present study investigates infiltration models in soil porous media, useful
tools for proactive and rehab actions, to safeguard surface and subsurface soft
water, which have become even more scarce. Integrated management plans of
water basins are of crucial importance since intensively cultivated areas are
adding huge quantities of fertilizers to the soil, affecting surface water basins and
groundwater.
Aquifers are progressively being nitrified on account of the nitrogen-based
fertilizer surplus, rendering water for human consumption not potable. Welltested
solute leaching models, standalone or part of a model package, provide
rapid site-specific estimates of the leaching potential of chemical agents, mostly
nitrates, below the root zone of crops and the impact of leaching toward
groundwater.
Most of the infiltration models examined, were process-based or
conceptual approaches. Nonetheless, empirical prediction models, though rather
simplistic and therefore not preferrable, demonstrate certain advantages, such as
less demanding extensive calibration database information requirements, which
in many cases are unavailable, not to mention a stochastic approach and the
involvement of artificial intelligence (AI). Models were categorized according
to the porous medium and agents to be monitored. Integrated packages of
nutrient models are irreplaceable elements for extensive catchments to monitor
the terrestrial nitrogen-balanced cycle and to contribute to policy making as
regards soft water management.
Infiltration is a complex environmental process by which an aquatic
solution i.e. rainwater, irrigation, potentially contaminated with
insoluble/dispersed particles, enters the ground under the gravity force or the
capillary action in deeper soil layers (percolation). Simple Hortonian
compression events (when overflow run exceeds land infiltration capacity and
depression storage capacity), silty-dominated topsoil with smooth texture characteristics and coarse-grained soil constituents are simplistically simulated
by lab columns, in different settlement modes, e.g. packed/non-packed. In the
present essay, suspended material, compression model equations were tested
over laboratory setup columns by optimizing model fitting curves via nonlinear
regression analysis, one of which received the lowest Standard Error of Estimate
value, among eight overall evaluated models, an indication of the best fitting
model performance of the ongoing soil-water suspension compression. By using
multiple linear regression, modifications were made of all eight tested models
which are presented herein. The modified models under examination, were
structured upon the linear dependence of their parameters on the selected
variables: silt-water concentration, inner cylinder diameter and effective
porosity. A modified Kang et al model, proposed to predict soil water
compression phenomenon, by setting coefficients of linear dependence on
independent variables. A higher applicability is to be achieved, due to the better
fitting of the tested field values, after the conclusion of a proper field calibration.
Practitioners would be able to implement the findings over certain field
engineering applications. The observed settlement response, during soil water
compression in packed/non packed column operating modes, is in compliance
with the relative bibliography.


