doktorska disertacija
Abstract
Onesnaženost voda velja za enega ključnih ekoloških problemov današnjega časa. V Sloveniji in
Sredozemlju je v rekah in morju poleg običajnih onesnaževal v visokih koncentracijah prisotno
tudi živo srebro. Kot del reševanja problema onesnaženosti voda pri nas in v svetu razvijamo
matematične modele, s katerimi je možno določiti obseg onesnaženja in možnosti sanacije.
Model PCFLOW3D je 3D baroklini matematični model, ki ponuja možnost modeliranja hidrodinami
čnih in nekaterih okoljskih spremenljivk v realnem času. V okviru disertacije smo model
PCFLOW3D nadgradili z dvema konceptualnima pelagičnima modeloma za nižji del prehranjevalne
verige, s konceptualnim bentoškim modelom mineralizacije organske snovi v sedimentih, z
empiričnim modelom valovanja za Tržaški zaliv in z empiričnimi modeli za določanje skalarnih
polj nekaterih okoljskih spremenljivk ter zvrsti živega srebra v Sredozemskem morju. Pri razvoju
konceptualnih modelov smo se opirali predvsem na deli Crispi in sod. (2002) in Fennel in sod.
(2006), pri razvoju empiričnih modelov pa smo uporabili metodi modelnih dreves (program
Weka) in ANFIS kombinacijo mehke logike in nevronskih mrež (program Matlab Fuzzy Logic
Toolbox).
Nadgrajeni model PCFLOW3D ponuja možnost kvantitativnega modeliranja fosfatov, nitratov,
amoniaka, kisika, zooplanktona, koncentracije klorofila-a in dveh velikostnih razredov fitoplanktona
in detrita. Med procesi v vodnem stolpcu upošteva mineralizacijo detrita, nitrifikacijo, agregacijo
manjših delcev v večje ter izmenjavo kisika med vodo in atmosfero. Bentoški model
je namenjen določanju vpliva procesov mineralizacije organske snovi v sedimentih (aerobne respiracije,
nitrifikacije in denitrifikacije) na vrednosti spremenljivk v vodnem stolpcu. Kombinirani
bentoško-pelagični model omogoča ekološko modeliranje odprtih in priobalnih morij ter celinskih
voda, z nekaj popravki pa ga lahko uporabljamo tudi za modeliranje kroženja ogljika ali drugih
elementov.
Empirični model valovanja za Tržaški zaliv ponuja možnost napovedi povprečne in vrčne periode
valov, povprečne in maksimalne višine valov, srednje smeri potovanja valov in razpršenosti valov
pri vršni periodi. Vhodne spremenljivke so pretekle vrednosti hitrosti in smeri vetra, sunkov
vetra, privetrišča ter hitrosti in smeri tokov v vrhnjem sloju morja; iz meteorološke napovedi
vhodnih spremenljivk lahko značilne vrednosti valov napovemo za poljubno število ur vnaprej.
Rezultati modela so primerljivi z rezultati fizikalnih modelov valovanja, če jih uporabimo v priobalnih
območjih s kompleksnimi hidrodinamičnimi in meteorološkimi pogoji. Model valovanja
lahko v prihodnje uporabimo za modeliranje vpliva valovanja na resuspenzijo sedimentov z dna
ter s tem povezanih sprememb koncentracije na delce vezanih onesnaževal v vodnem stolpcu in
v sedimentih Tržaškega zaliva.
Empirični modeli za določanje skalarnih polj konduktivnosti, transmisije in kisika v Sredozemskem
morju so zelo dobro ocenjeni (r t 1.0); primarno so namenjeni natančnejšemu določanju
koncentracije zvrsti živega srebra, ki je od njih odvisna. Empirični modeli za določanje skalarnih
polj celokupnega, elementarnega, reaktivnega, monometil in dimetil živega srebra v Sredozemskem
morju so slabˇse ocenjeni, iz strukture modelov pa lahko zaključimo, da na koncentracijo
zvrsti živega srebra v Sredozemskem morju najbolj vplivajo koncentracija kisika, prisotnost delcev
organske snovi in intenziteta svetlobe. Podobni empiriˇcni modeli za določanje koncentracije
živega srebra v vodnem okolju, ki jih v prihodnje lahko razvijemo z orodji mehkega računanja
ali z drugimi naprednejšimi metodami umetne inteligence, ponujajo v kombinaciji s konceptualnim
bentoško-pelagičnim modelom in ostalimi moduli modela PCFLOW3D možnost določitve
skalarnih polj zvrsti živega srebra v Sredozemskem morju ali na drugih območjih s podobno
problematiko.
Keywords
gradbeništvo;disertacije;oceanografija;ekološko modeliranje;živo srebro;nižji del prehranjevalne verige;valovi;matematični model PCFLOW3D;Sredozemsko morje;Tržaški zaliv;
Data
Language: |
Slovenian |
Year of publishing: |
2011 |
Typology: |
2.08 - Doctoral Dissertation |
Organization: |
UL FGG - Faculty of Civil and Geodetic Engineering |
Publisher: |
[N. Sirnik] |
UDC: |
519.87:504.054:532.05(262.3)(043.3) |
COBISS: |
5433953
|
Views: |
2804 |
Downloads: |
451 |
Average score: |
0 (0 votes) |
Metadata: |
|
Other data
Secondary language: |
English |
Secondary title: |
Modelling the hydrodinamic, oceanographic and environmental variables in marine environment with emphasis on mercury concentrations |
Secondary abstract: |
Today, water pollution is considered to be one of the major ecological problems. In Slovenia and
the Mediterranean, mercury, among other pollutants, is present in high concentrations. One
of the ways to tackle the problem of water pollution is to use mathematical models that can
help in determining the level of pollution and the possible ways of sanitation. The PCFLOW3D
model is a 3D baroclinic mathematical model, which enables us to model hydrodynamic and
some environmental variables in real time. In the dissertation, the PCFLOW3D model was upgraded
with two conceptual pelagic models for the lower trophic level, with a conceptual benthic
model of organic matter mineralization in sediments, with an empirical wave model for the Gulf
of Trieste and with empirical models for scalar fields of certain environmental variables and of
the concentrations of mercury species in the Mediterranean region. The two lower trophic level
models were based primarily on Crispi et al. (2002) and Fennel et al. (2006). The empirical
models were generated using two artificial intelligence tools, namely model trees (Weka software)
and the ANFIS combination of fuzzy systems and neural networks (Matlab Fuzzy Logic Toolbox
software).
With the upgraded PCFLOW3D model it is possible to quantitatively model the concentration
of phosphates, nitrates, ammonium, dissolved oxygen, zooplankton, chlorophyll-a and two size
classes of phytoplankton and detritus. Among the processes in the water column it takes into
account the mineralization of detritus, nitrification, aggregation of smaller organic particles into
larger ones and the exchange of oxygen between water and atmosphere. The benthic model is
designed to determine the impact of aerobic mineralization, nitrification, and denitrification in
sediments on the concentration of environmental variables in the water column. The combined
benthic-pelagic model enables us to simulate the biogeochemical cycling of the listed variables in
open oceans, on continental margins and in continental waters, and, with only little additional
upgrade, it can be used to model the carbon cycle or the cycle of other important elements or
complexes.
The empirical wave model for the Gulf of Trieste forecasts average and peak wave period, average
and maximum wave height, mean direction of wave propagation and spread at peak period.
The input variables are: past values of wind speed and direction, fetch, speed, and direction of
currents in the top water layer; from the meteorological forecast of the input variables waves
can be forecasted even 72 hours ahead. The results of modelling are comparable with those of
physical wind-wave models, if used in near shore sites with complex hydrodynamic and meteorological
conditions; above all the forecast of wave height can be used well with r t 0.95. In
future, the model can be used to quantify the impact of waves on the resuspension of sediments
and the consecutive change in concentration of mercury species in the water column of the Gulf
of Trieste.
Empirical models for scalar fields of conductivity, transmission and dissolved oxygen concentration
in the Mediterranean Sea are very well correlated with measurements (r t 1.0). These
models can be used for a more accurate determination of concentrations of mercury species in the
water column which depends on them. The empirical models for scalar fields of total, elementary,
reactive, monomethyl and dimethyl mercury in the Mediterranean Sea are estimated as worse,
but from the structure of the model trees it can be concluded that the concentration of mercury
species in the Mediterranean Sea is most influenced by the concentration of dissolved oxygen,
the concentration of organic matter and light intensity. In future, similar empirical models for
the determination of mercury species concentration in water compartments can be generated
with other artificial intelligence tools, among which the methods which combine induction and
deduction seem most suitable. The combination of such models and the benthic-pelagic conceptual
PCFLOW3D model can be used to determine the scalar fields of mercury species in the
Mediterranean Sea or in other water compartments with similar pollution problems. |
Secondary keywords: |
civil engineering;doctoral thesis;oceanography;ecological modelling;mercury;lower trophic level;waves;mathematical model PCFLOW3D;Mediterranean Sea;the Gulf of Trieste;Živo srebro;Disertacije; |
URN: |
URN:NBN:SI |
File type: |
application/pdf |
Type (COBISS): |
Dissertation |
Thesis comment: |
Univ. Ljubljana, Fak. za gradbeništvo in geodezijo |
Pages: |
XXXII, 309 str. |
ID: |
1725259 |