Avtor | Leto | Naslov | Mentor |
ZABRET Katarina | 2018 | Vpliv meteoroloških in vegetacijskih parametrov na prestrezanje padavin | izr. prof. dr. Mojca Šraj prof. dr. Jože Rakovec |
BEZAK Nejc | 2016 | Povezanost fluvialnega premeščanja suspendiranih snovi z drugimi hidrološkimi procesi | prof. dr. Matjaž Mikoš doc. dr. Mojca Šraj |
LAMOVEC Peter | 2013 | Določitev poplavljenih območij z uporabo daljinskega zaznavanja in strojnega učenja | izr. prof. dr. Krištof Oštir, prof. dr. Matjaž Mikoš |
PETAN Sašo | 2010 | Meritve in modeliranje erozivnosti padavin kot parametra erozije tal | prof. dr. Matjaž Mikoš, prof. dr. Jože Rakovec |
KRYŽANOWSKI Andrej | 2009 | Abrazijska odpornost betonskih površin na vodnih zgradbah |
prof. dr. Matjaž Mikoš, izr. prof. dr. Igor Planinc |
RUSJAN Simon | 2008 | Hidrološke kontrole sproščanja hranil v porečjih |
prof. dr. Matjaž Mikoš, prof. dr. Mitja Brilly |
ZORN Matija | 2007 | Recentni geomorfni procesi na rečno-denudacijskem reliefu na primeru porečja Dragonje | izr.
prof. dr. K. Natek, prof. dr. M. Mikoš |
KOBOLD Mira | 2007 | Vpliv napake ocene padavin na napako napovedi odtoka pri napovedovanju poplav |
prof. dr. M. Brilly, prof. dr. J. Rakovec |
KRZYK Mario | 2004 | Dvodimenzijsko matematično modeliranje toka v strmih strugah |
izr. prof. dr. M. Četina,
izr. prof. dr. M. Mikoš |
BIZJAK Aleš | 2003 | Sintezni postopek ocenjevanja hidromorfološkega stanja rečnih koridorjev, razvit z analizo stanja na reki Dragonji |
izr. prof. dr. M. Mikoš,
prof. dr. M. Brilly |
ŠRAJ Mojca | 2003 | Modeliranje in merjenje prestreženih padavin |
prof.dr. M. Brilly,
izr.prof.dr. M. Mikoš |
PETKOVŠEK Gregor | 2002 | Kvantifikacija in modeliranje erozije tal z aplikacijo na povodju Dragonje |
izr. prof. dr. M. Mikoš,
izr. prof. dr. M. Četina |
GLOBEVNIK Lidija | 2001 | Celosten pristop k urejanju voda v povodjih |
prof. dr. M. Brilly,
izr. prof. dr. M. Mikoš |
SMOLAR ŽVANUT Nataša | 2000 | Vloga perifitonskih alg pri določanju ekološko sprejemljivega pretoka v tekočih vodah |
prof. dr. D. Vrhovšek,
izr. prof. dr. M. Mikoš |
GORIŠEK Metka | 1995 | Difuzijsko gibanje snovi raztopljenih v vodi skozi porozni prostor iz drobnozrnatih materialov | prof. dr. M. Brilly |
KRYŽANOWSKI Andrej:
Abrazijska odpornost betonskih površin na vodnih zgradbah
Mentor:
prof. dr. Matjaž
Mikoš
Somentor: izr. prof. dr.
Igor Planinc
zagovor: julij 2009
Povzetek
V disertaciji je
obravnavana problematika abrazijske erozije betonov, ki se na vodnih zgradbah
pojavlja kot posledica abrazijskega delovanja rinjenih plavin. Ključni problem
modeliranja abrazijske erozije v laboratoriju predstavlja težavnost modeliranja
ustreznih hidravličnih pogojev za nastanek procesa in razmer ekvivalentnih
tistim v naravi ter interpretacija rezultatov meritev, ki praviloma bazira zgolj
na kvalitativnih analizah. Validacija rezultatov laboratorijskih raziskav je
tako možna samo z izvedbo preskusa v normalnih obratovalnih razmerah v naravi. V
prvem delu disertacije smo poskušali razjasniti delovanje vodnega toka z
rinjenimi plavinami na izpostavljene dele betonskih površin na vodnih zgradbah
in določiti primerne sestave betonov pri čemer smo želeli doseči predvsem
ustrezne trdnostne karakteristike preskusnih betonov in njihovo izvedljivost –
to je možnost vgradnje izbranih materialov brez posebnih predpriprav. Ker je
proces abrazijske erozije sam po sebi izredno kompleksen, smo se v okviru našega
dela omejili na raziskavo abrazijske odpornosti betonov z modifikacijami osnovne
sestave betonov, vgrajenih na evakuacijskih objektih hidroelektrarn na Savi. V
šest polj smo, poleg betona z osnovno sestavo betona, vgradili beton z dodatki
polimernega veziva, jeklenih vlaken, poli-propilenskih vlaken ter beton z
gumenim agregatom. V drugem delu smo primerjali analizo rezultatov
laboratorijskih raziskav in meritve opravljene na poskusnih poljih v naravi. Pri
analizi abrazijske odpornosti betonov na vodnih zgradbah je bistvenega pomena
ustrezna interpretacija dobljenih rezultatov, ki se je do sedaj največkrat
omejevala na primerjalno analizo rezultatov raziskav abrazijske odpornosti,
opravljenih po standardnih postopkih v laboratoriju. Abrazijska odpornost je
bila analizirana z metodo podvodne abrazije po postopku
ASTM C 1138, dobljene rezultate pa
smo primerjali z meritvami opravljenimi v naravi na poskusnih poljih izvedenih v
podslapju hidroelektrarne Vrhovo – potrjena je bila soodvisnost rezultatov in
meritev. V zaključkih
v disertaciji ugotavljamo, da lahko, na osnovi meritev obrabe betonov po
postopku ASTM C 1138
in ob upoštevanju podobnih hidroloških pogojev ter obratovalnih razmer, ki so
veljali v času izvajanja disertacije, napovemo časovni razvoj obrabe betonov na
vodnih zgradbah. Rezultati potrjujejo ustreznost metode
ASTM C 1138
za napoved razvoja abrazijskih procesov in dinamike obrabe betonskih površin na
jezovnih zgradbah elektrarn na Savi.
Ključne besede:
hidravlika,
hidrotehnični objekti, abrazija, abrazijske poškodbe, abrazijska odpornost,
eksperiment, beton
UDK:
627.2/.8:691.32(043.2)
Abstract
The PhD thesis deals with
the abrasive wear of concrete occurring in hydraulic structures as a consequence
of the abrasive action of waterborne particles. The key problem of modelling
abrasive wear under laboratory conditions lies in the difficulty of modelling
the proper hydraulic conditions for triggering the process and conditions
equivalent to those in the field, as well as the in the interpretation of
measuring results, which is, as a rule, based on qualitative analysis only. The
validation of the results obtained in the laboratory is therefore possible only
by performing tests under normal operational conditions in the field. In the
first part of the thesis we tried to define the action of waterborne solid
particles to exposed concrete surfaces in hydraulic structures and define the
appropriate concrete mixtures, thus trying to obtain the required strength
characteristics of the test concretes and their feasibility – the possibility of
building-in of the chosen materials without any special preconditioning. Since
the process of abrasive wear in itself is a highly complex one, we focused on
studying the abrasion resistance of concrete with modifications of the basic
composites, built in the evacuation structures of the hydropower plants on the
Sava River, Slovenia. The basic composition of concretes built into the six
plots was modified by polymer additives, steel fibers, polypropylene fibers, and
granular rubber. In the second part we compared the analysis of the results
obtained in laboratory simulations and the measurements performed in the test
plots in the field. When analysing abrasion resistance of concrete in hydraulic
structures the interpretation of the results obtained is of major importance;
until now this was usually limited to the comparative analysis of the results of
studies on abrasion resistance, which were performed in the laboratory using the
standard procedures. Abrasion resistance was investigated using the standard
ASTM C 1138 test method of underwater
abrasion, and the results were then compared to the measurements performed in
the test fields of the stilling basin of the Vrhovo hydropower plant (HPP) – the
correlation of the results and measurements was confirmed.
In the conclusion, we argue
that based on the measurements of concrete wear following the standard
ASTM C 1138 and considering similar
hydrological and operational conditions at the time of performing the thesis
research, a time machine of concrete wear in hydraulic structures can be
predicted. The results confirm the adequacy of the
ASTM C 1138 method for predicting the
advancement of the abrasion processes and the dynamics of wear of concrete
surfaces in the dam structures of the Sava River HPPs.
Key words:
hydraulics,
hydraulical structures, abrasive action, abrasion damage, abrasive resistance,
test, concrete
UDC:
627.2/.8:691.32(043.2)
ZORN Matija:
Recentni geomorfni procesi na
rečno-denudacijskem reliefu na primeru porečja Dragonje
Mentor:
izr. prof. dr.
Karel Natek
Somentor:
prof. dr. Matjaž
Mikoš
zagovor: november 2007
Povzetek
Doktorsko delo je
prispevek k boljšemu razumevanju geomorfnih procesov v flišnih pokrajinah
jugozahodne Slovenije. Predstavljene so meritve erozijsko-denudacijskih procesov
v porečju Dragonje. Meritve so potekale od februarja 2005 do maja 2006. Merili
smo površinsko spiranje na treh različnih rabah tal, umikanje strmih golih
flišnih pobočij, premikanje gradiva po erozijskem jarku in kemično denudacijo v
porečju. V manjšem obsegu smo merili še žlebično in vetrno erozijo. Meritve so
bile tedenske, rezultati pa so predstavljeni tedensko, mesečno ter po letnih
časih in so korelirani z vremenskimi razmerami (padavinami, temperaturami in
vetrom).
V porečju je najhitrejše umikanje
strmih flišnih pobočij, ki se umikajo skoraj s hitrostjo 5 cm/leto. Premikanje
gradiva po erozijskem jarku je skoraj šestkrat počasnejše od sproščanja fliša.
Skoraj šestkrat počasnejša pa je tudi hitrost zniževanja površja zaradi
površinskega spiranja na goli prsti (~8,5 mm/leto). Zniževanje površja zaradi
površinskega spiranja v gozdu (~0,4 mm/leto) je za faktor 130 počasnejše od
umikanja strmih pobočij, zniževanje površja na travniku pa je počasnejše za
faktor 300 (0,16 mm/leto). Najpočasnejša je kemična denudacija (0,07 mm/leto),
ki je kar za faktor 744 počasnejša od umikanja strmih flišnih pobočij.
Ključne besede:
geografija,
geomorfologija, erozijski procesi, erozija prsti, umikanje pobočij, fliš,
rečno-denudacijski relief, Istra, Slovenija
UDK:
911.2:551.435(497.4-14)(043.3)
Abstract
The
aim of this doctoral thesis is better understanding of geomorphic processes in
flysch landscapes in south-western Slovenia. Measurements of erosion and
denudation processes in the Dragonja river basin are presented. They took place
weekly between February 2005 and May 2006. We measured (1) interrill erosion on
three different land uses, (2) rockwall retreat on steep flysch slopes, (3)
movements of debris in erosion gully, (4) chemical denudation and in smaller
extend also (5) rill erosion and (6) wind erosion. The results are presented
according to weeks, months and seasons and are correlated with weather
conditions of that time (precipitation, temperature and wind).
Rockwall retreat on steep flysch slopes, with speed of almost 5 cm/year was the
fastest among all measured erosion processes. The slope processes in erosion
gully were almost six times slower. Surface lowering due to interrill erosion on
bare flysch soil has approximately the same speed as processes in erosion gully
(~8.5 mm/year). Surface lowering due to interrill erosion in the forest (~0.4
mm/year) is almost 130 times slower in relation to rockwall retreat and on
meadows (0.16 mm/year) 300 times slower in relation to rockwall retreat.
Chemical denudation is the slowest process (0.07 mm/year), being 744 times
slower than rockwall retreat.
Key words:
geography,
geomorphology, erosion processes, soil erosion, rockwall retreat, flysch,
fluvial-denudational relief, Istria, Slovenia
UDC:
911.2:551.435(497.4-14)(043.3)
KOBOLD Mira : Vpliv napake ocene padavin na napako napovedi
odtoka pri napovedovanju poplav
Mentor: prof. dr. Mitja Brilly
Somentor: prof. dr. Jože Rakovec
zagovor: april 2007
Povzetek
Doktorska
disertacija obravnava vpliv padavin, vlažnosti tal in evapotranspiracije na
oceno odtoka s povodja Savinje. Za izračun odtoka kot tudi ostalih hidroloških
spremenljivk je bil uporabljen švedski konceptualni model HBV, verzija HBV-96.
Model smo v letu 2003 v okviru evropskega raziskovalnega projekta “European
Flood Forecasting System”, ki je tekel v petem okvirnem programu Evropske
skupnosti, prvič implementirali v Sloveniji na povodju Savinje.
Prikazani so rezultati umerjanja dveh modelov povodja Savinje
z različnim časovnim korakom, dan in ura. Za simulacijo odtoka in napoved
visokih voda in poplav je primeren model s časovnim korakom ene ure, vendar
število petih ombrografskih postaj na povodju in okolici ni dovolj za točno
oceno ploskovnih padavin in zadovoljivo umerjanje modela, saj ima topografija
povodja močan vpliv na meteorološko dogajanje v porečju, zlasti v zgornjem toku
Savinje. Povprečna relativna napaka pretoka znaša okrog 40 % za celotno območje
pretokov, za visokovodne konice pa okrog 25 %.
Splošni problem mreže padavinskih postaj je njena neustreznost za hidrološke
aplikacije. Z analizo vpliva števila padavinskih postaj na oceno ploskovne
količine padavin smo pokazali, da lahko število postaj odločilno vpliva na oceno
količine padavin na povodju in s tem na odstopanje simuliranega odtoka od
opazovanj. Analize kažejo zlasti veliko odstopanje pri majhni količini padavin
na povodju, ko prevladuje nevihtni tip padavin in so padavine običajno lokalno
omejene. Napaka pri majhni količini padavin na povodju lahko preseže 100 %. Pri
padavinah, ki povzročijo visoke vode in poplave, je relativna napaka manjša, do
50 %. S korigiranjem padavin v modelu HBV smo za nekaj padavinskih dogodkov
pokazali, da je ob pravilnem opisu padavin v modelu ujemanje simuliranega
pretoka z merjenim dobro. Z analizo prostorske spremenljivosti padavin na
povodju Savinje smo izdelali predlog optimalnega števila postaj na povodju z
zveznim beleženjem intenzitete padavin za potrebe hidrološkega modeliranja.
Ker nezadostna mreža ombrografskih postaj na povodju Savinje
ne da pravilne ocene ploskovnih padavin, smo v nalogi obravnavali uporabnost
radarskih padavin v hidrološkem modeliranju. Analiza dveh padavinskih dogodkov
je pokazala, da so trenutne urne akumulacije padavin iz radarskega centra na
Lisci neuporabne za modele padavine-odtok, saj so le te podcenjene za faktor 3
do faktor 6.
Izvedli smo analizo občutljivosti konceptualnih modelov na
padavine kot osnovni vhodni podatek v model. Izpeljali smo odvisnost med napako
v padavinah in odstopanjem simulirane visokovodne konice zaradi te napake.
Odvisnost je polinom drugega reda in ni odvisna od velikosti povodja. Z
izpeljano enačbo lahko ovrednotimo odstopanje pretoka konice visokovodnega vala,
če vemo, kolikšna je napaka v oceni ploskovne količine padavin. Razmerje med
napako pretoka konice visokovodnega vala in napako padavin je pri podcenjenih
padavinah pod 1.6, pri precenjenih padavinah pa nad to vrednostjo. Pri
precenitvi padavin za 50 % se napaka v odtoku že skoraj podvoji.
Z analizo poplav iz leta 1980 in leta 1990 smo za povodje
Savinje proučili vpliv vlažnosti tal in evapotranspiracije na odtok. V obeh
poplavnih dogodkih je bila razporeditev in količina padavin na povodju podobna,
pretok poplavnega vala leta 1980 pa za okrog 200 m3/s manjši.
Različne začetne hidrološke razmere so povzročile, da je ob nizki predhodni
vlažnosti pred poplavnim dogodkom leta 1980 infiltracija v tla predstavljala kar
37 % bruto padavin, pri poplavi leta 1990 pa le 13 %. Evapotranspiracija na
poplavne dogodke nima velikega vpliva, v obravnavanih dogodkih je predstavljala
le 1 % bruto padavin. Z modelom HBV je bila narejena simulacija poplavnega
hidrograma za november 1990 s padavinami iz poplavnega dogodka oktobra 1980, ko
je bilo padavin na povodju za 27 mm več kot v dogodku leta 1990. Izračun je
pokazal, da bi pretok s temi padavinami dosegel skoraj tisočletno povratno dobo
velikih pretokov.
V zaključnem delu naloge so prikazani rezultati simulacij konceptualnih modelov
HEC-1 in HBV in primernost teh dveh modelov za napovedovanje hudourniških poplav
s prednostmi in pomanjkljivostmi vsakega od modelov. Izvedli smo še primerjalno
analizo vseh obstoječih modelov na povodju Savinje ter podali oceno primernosti
za napovedovanje visokih voda in poplav. Prikazane in analizirane so situacije
treh padavinskih dogodkov s prognoziranimi padavinami modela ALADIN/SI, med
njimi poplava iz novembra 1998. Operativna uporabnost hidroloških napovedi s
produkti modela ALADIN/SI je prikazana z rezultati polletnih testnih zagonov
modela HBV.
Ključne besede:
poplave, padavine, odtok, Savinja, modeliranje, HBV, HEC-1, vlažnost tal,
evapotranspiracija, napovedovanje, ALADIN/SI
UDK:
627.51:627.131:551.501:519.87(043.3)
Abstract
The dissertation treats the
influence of precipitation, soil moisture and evapotranspiration on the
estimation of runoff from the Savinja catchment. The Swedish conceptual HBV-96
model was used for runoff simulation as well for other hydrological variables.
The model was first applied in Slovenia in the year 2003 on the Savinja
catchment, in the frame of European research project “European Flood Forecasting
System” running under the fifth framework programme of the European Union.
The results of calibration of two models for the Savinja
catchment with two different time steps, day and hour, were presented in the
dissertation. The model with time step of one hour is suitable for flash flood
forecasting, but five recording raingauges on the catchment and its closeness is
not sufficient for accurate estimation of areal precipitation and satisfactory
calibration of the model because the topography of the catchment has strong
influence on the meteorological happening of the catchment, especially in the
upper stream of the Savinja. The average relative error of discharge is about 40
% for the entire range of discharges and 25 % for flood peaks.
General problem of precipitation network is its unsuitability
for hydrological applications. The analysis of number of raingauges on the areal
estimation of precipitation has shown that number of raingauges is crucial for
the estimation of areal precipitation, what leads to the deviation of simulated
runoff from observed values. Analyses show great deviations especially by small
amount of precipitation on the catchment when there is storm and usually local
precipitation. The relative error can exceed 100 % in the case of small amount
of precipitation. The relative error is smaller by precipitation causing high
waters and floods and can reach 50 %. The correction of precipitation in the HBV
model for some high water events showed that simulated runoff fits with measured
one well by proper precipitation input. Analysis of spatial distribution of
precipitation on the Savinja catchment gave the proposal of optimal number of
recording raingauges on the catchment.
Insufficient number of precipitation gages does not give the
accurate estimation of areal precipitation, therefore the applicability of radar
precipitation was treated as input in hydrological modelling. The analysis of
two precipitation events has shown that hourly radar accumulations of
precipitation from radar centre on Mount Lisca are underestimated for factors
between 3 and 6 and thus useless for the rainfall-runoff models.
The sensitivity analysis of conceptual models on
precipitation input was investigated. The relationship between an error in
precipitation amount and deviation of simulated peak discharge was carried out.
The equation is quadratic polynomial and the relationship is not dependent on
the catchment area. Derived equation can be used to asses the deviation of
predicted runoff in the case of known deviation in the estimation of areal
precipitation. The ratio of errors between peak discharges and underestimated
precipitation is under 1.6 and above this value by overestimated precipitation.
The error in runoff is almost doubled by 50 % of overestimated precipitation.
With analysis of floods in 1980 and 1990 on the Savinja
catchment the impacts of soil moisture and evapotranspiration on runoff were
investigated. In both flood events, the distribution and amount of precipitation
were similar on the catchment, but runoff in the year 1980 was about 200 m3/s
lower compared to the year 1990. The differences in initial hydrological
circumstances caused that 37 % of total precipitation was taken by infiltration
into the soil in flood of 1980 and only 13 % in flood of 1990.
Evapotranspiration do not have large impact on flood waves, in analysed events
the evapotranspiration was only 1 % of total precipitation. Flood hydrograph in
November 1990 was simulated by HBV model with precipitation which caused flood
in October 1980, which was greater by 27 mm on the catchment. The results of
calculation showed that the peak discharge would reach the thousand year return
period flood in that case.
In the final part of the dissertation the results of simulations of conceptual
models HEC-1 and HBV were presented. The advantages and disadvantages of both
models and their suitability for flash flood forecasting are given. The
intercomparison of all models covering the Savinja catchment was carried out.
The assessment of suitability of models for flood forecasting is given. Three
precipitation events are presented and analysed with predicted precipitation of
ALADIN/SI model, one of them is flood from November 1998. Operational
applicability of hydrological predictions by products of ALADIN/SI model are
presented with the results of half-yearly test runs of the HBV model.
Key words:
floods, rainfall,
runoff, Savinja, modelling, HBV, HEC-1, soil moisture, evapotranspiration,
forecasting, ALADIN/SI
UDC:
627.51:627.131:551.501:519.87(043.3)
BIZJAK Aleš :
Sintezni postopek ocenjevanja
hidromorfološkega stanja rečnih koridorjev, razvit z analizo stanja na reki
Dragonji
Mentor: izr. prof. dr. Matjaž Mikoš
Somentor: prof. dr. Mitja Brilly
zagovor: november 2003
Povzetek
V raziskavi obravnavamo razvoj nove
sintezne metode za oceno hidromorfološkega stanja rečnih koridorjev, izdelane z
analizo stanja na reki Dragonji. Izhodišče izdelave nove metode je bila hitrost
zajema podatkov in natančnost metode. Pred izvedbo aplikativnega dela raziskave
smo zasnovali koncept hitrega transektnega zajema podatkov ter izdelali obsežen
hidromorfološki inventarizacijski list s kombiniranim naborom hidromorfoloških
spremenljivk v tujini najpogosteje uporabljanih metod, švedske Riparian, Channel
and Environmental Inventory, ameriške Stream Visual Assessment Protocol,
angleške River Habitat Survey in nemške Gewässerstrukturgütebewertung.
Terenski zajem podatkov smo izvedli na osnovi
hidromorfološkega inventarizacijskega lista v 288 predhodno kartografsko
določenih transektih vzdolž teka reke Dragonje na medosni razdalji 100 m in
izdelali obsežno podatkovno bazo 148 hidromorfoloških spremenljivk v vsakem
izmed transektov.
Podatkovno bazo smo za potrebe izdelave sintezne metode
uredili v bazo 99 diskretnih hidromorfoloških spremenljivk. Z uporabo
statističnih orodij (analiza glavnih komponent, multivariatna analiza variance,
korelacija) in orodij strojnega učenja (odločitveno drevo) smo jo sistematično
reducirali v nabor 52 najpomembnejših hidromorfoloških spremenljivk za oceno
hidromorfološkega stanja. Na osnovi pregleda in analize izbranih
hidromorfoloških spremenljivk smo določili končni nabor 35 hidromorfoloških
spremenljivk in izdelali sintezno metodo za oceno hidromorfološkega stanja
rečnih koridorjev. Strukturo in ocenjevalni sistem metode smo prilagodili
transektnemu zajemu podatkov in metodo aplicirali v 288 transektih vzdolž reke
Dragonje.
Na osnovi predhodnih rezultatov testiranja omenjenih metod na
petih testnih odsekih reke Dragonje in petih testnih odsekih reke Reke smo kot
primerjalno metodo v naši raziskavi izbrali nemško metodo
Gewässerstrukturgütebewertung (GSGB), ki za izdelavo hidromorfološke ocene
potrebuje 32 hidromorfoloških spremenljivk. Metodo smo z originalnim konceptom
odsečnega zajema v 111 odsekih in prirejenim transektnim zajemom podatkov v 288
transektih aplicirali vzdolž celotnega teka reke Dragonje.
S primerjalno analizo ocen odsečnega zajema po nemški metodi
in transektnega zajema po nemški in sintezni metodi smo ugotovili, da lahko na
primeru reke Dragonje za izdelavo ocene hidromorfološkega stanja rečnega
koridorja namesto časovno zahtevnega odsečnega zajema uporabimo hitrejši
transektni zajem podatkov, kljub temu da na reki Dragonji skupna dolžina
transektov zajame le 20 % skupne dolžine odsekov. Prav tako smo ugotovili, da
lahko ocenjevalni sistem transektne sintezne metode uspešno apliciramo tudi v
daljših odsekih reke Dragonje.
Analiza vpliva subjektivnega faktorja, izdelana na osnovi 144 terenskih meritev
predhodno šolanih študentov je pokazala, da se v večini testnih transektov glede
na izračunane vrednosti t-statistike rezultati študentskih meritev v obeh
metodah razlikujejo od rezultatov ekspertnih meritev. Pregled odstopanj
vrednosti študentskih meritev od ekspertnih za posamezne spremenjlivke obeh
metod je pokazal, da bi bilo potrebno popisovalce hidromorfoloških kakovosti
bolj poglobljeno uvajati v tovrstno delo v rečnih koridorjih. Izračun intervalov
zaupanja za študentske meritve za obe metodi kaže, da so študentje z manj
raztrosa in zanesljiveje določali hidromorfološke kakovostne razrede testnih
transektov po sintezni metodi. Primerjalna analiza rezultatov meritev študentov
z izračunom H- statistike neparametrskega testa Kruskal - Wallis pa je pokazala,
da so študentje tako po metodi GSGB kakor po sintezni metodi določevali podobne
hidromorfološke ocene in jih tudi podobno razvrščali v hidromorfološke
kakovostne razrede.
Ključne besede:
Dragonja, ekohidrologija, hidromorfološko stanje, krajinska ekologija, rečni
koridor, terenski poskus, transektni zajem podatkov, urejanje vodotokov
UDK:
504.4.05 + 556.18 + 556.52 / .53 +
556.537 (043.3)
Abstract
The research deals with the
development of a new synthesis method of assessing the hydromorphological status
of river corridors, based on state analysis of the Dragonja River. The bases of
the new method were time effectiveness of data gathering and accuracy of the
method. Prior to the practical research implementation, a concept of a rapid
transect data gathering was designed. Additionally, an extensive
hydromorphological record sheet was elaborated, including a combined list of
hydromorphological variables of the methods most frequently used abroad: the
Swedish Riparian, Channel and Environmental Inventory, the American Stream
Visual Assessment Protocol, the British River Habitat Survey and the German
Gewässerstrukturgütebewertung.
On the basis of the hydromorphological record sheet, field
data gathering was carried out in 288 transects a priori cartographically
determined along the Dragonja water course at an axis distance of 100 m. In each
of the transects, an extensive data base of 148 hydromorphological variables was
designed.
In order to work out the synthesis method, a data base of 99
discreet hydromorphological variables was prepared. By way of a systematical
reduction, a list of 52 most significant hydromorphological variables, according
to the results of statistical (Principal Component Analysis, MANOVA,
Correlation) and machine learning tools (decision trees) was made. On the basis
of reviewing and analysis of the determined hydromorphological variables, a
final list of 35 hydromorphological variables was designed. Based on the final
list, the synthesis method of assessing the hydromorphological status of river
corridors was made. The structure and the assessment system of the method was
adapted to the transect data gathering. The method was applied in the 288
transects along the Dragonja River.
On the basis of the results of applying the mentioned methods
in five test reaches of the Dragonja River and five test reaches of the Reka
River the German Gewässerstrukturgütebewertung (GSGB) was chosen as the
comparison method of the research. For the hydromorphological assessment, this
method requires 32 hydromorphological variables. The method was applied along
the Dragonja River in two ways, by way of its original concept of data gathering
in 111 reaches, and by way of the adapted data gathering in 288 transects.
The comparison analysis of assessing the data gathering in
reaches according to the German method and the transect data gathering according
to the German and the synthesis method has shown that in the case of the
Dragonja River, the faster transect data gathering can be used instead of the
more time consuming data gathering in the river reaches, despite the fact that
the sum of the transect lenghts amounts to only 20 % of the total lenght of the
reaches. Additionally, it was established that it is possible to successfully
apply the assessment system of the transect synthesis method even to reaches of
the Dragonja river.
Analysis of the factor of subjectiveness, conducted on the basis of 144 field
measurements of pre-trained students has shown that according to the calculated
values of the t-statistics in the majority of the test transects the results of
student measurements differ from expert measurements in both methods. A review
of deviations of measurements made by students in comparison to the expert
measurements has indicated that assessors of hydromorphological qualities should
receive a more in-depth introduction into the respective work in river
corridors. The calculated confidence intervals for student measurements
regarding both methods shows that the students had used the synthesis method
with less variance and with a higher level of accuracy determined the
hydromorphological quality classes of the transects. A comparison analysis of
the results of the student measurements by calculating the H-statistics of the
non-parametric Kruskal - Wallis test has shown that the students determined the
hydromorphological values as well as hydromorphological quality classes with a
high level of similarity in both methods.
Key words:
the Dragonja River,
ecohydrology, hydromorphological status, landscape ecology, river corridor,
field experiment, transect data gathering, river engineering
UDC:
504.4.05 + 556.18 + 556.52 / .53 +
556.537 (043.3)
ŠRAJ Mojca
:
Modeliranje in merjenje prestreženih padavin
Mentor: prof. dr. Mitja Brilly
Somentor: izr. prof. dr. Matjaž Mikoš
zagovor: marec 2003
Povzetek
Doktorska
disertacija je del obsežnega znanstveno-raziskovalnega projekta “Dragonja:
Forest – Water – Soil – Climate Interactions”, ki ga od leta 1999 izvajata
univerza Vrije Universiteit iz Amsterdama in Univerza v Ljubljani. Za
eksperimentalno povodje je bilo izbrano povodje Dragonje, ki je zanimivo zaradi
intenzivnih naravnih procesov zaraščanja površin z gozdom v zadnjih desetletjih,
kar je pripeljalo do zmanjšanja pretokov, tako minimalnih kot maksimalnih. Ob
tem opaznih padavinskih in temperaturnih sprememb ni bilo zaznati. Glavni namen
projekta kot celote je torej ugotoviti vpliv zaraščanja na vodno bilanco
celotnega povodja.
V okviru disertacije so bile narejene natančne meritve in
analize posameznih komponent hidrološkega kroga gozda s sodobno mersko opremo,
ter različni modeli prestreženih padavin na eksperimentalnem povodju Dragonje. V
ta namen sta bili v sodelovanju z univerzo Vrije Universiteit iz Amsterdama,
konec leta 1999 v listnatem gozdu na povodju Dragonje izbrani dve merski
ploskvi, ena na severnem (1419 m2) in druga na južnem pobočju hriba
(615 m2) nad sotočjem Dragonje in Rokave. Meritve so se začele v
jeseni 2000. V disertaciji je obravnavano obdobje enega leta od oktobra 2000 do
septembra 2001. Na obeh raziskovalnih ploskvah so se merile količine padavin nad
krošnjami, količine padavin, ki padejo skozi odprtine med krošnjami oz. kasneje
prikapljajo z njih ter delež padavin, ki odtečejo po deblu do tal. Vse naštete
količine so se merile avtomatsko z digitalnim zapisovanjem rezultatov na vsakih
10 minut. Padavine nad krošnjami so se merile z ombrografom, za kontrolo pa še s
totalizatorji z ročnim praznjenjem. Delež padavin, ki padejo skozi odprtine med
krošnjami in listi oz. kasneje prikapljajo z njih, se je meril s po dvema
kovinskima žlebovoma v kombinaciji s po desetimi premičnimi totalizatorji na
vsaki raziskovalni ploskvi, ki se jih je ročno praznilo in se jim vsakič
zamenjalo merilno mesto. Na ta način je bila zajeta tudi prostorska
spremenljivost tega deleža padavin. Odtok po deblu se je meril na dveh najbolj
tipičnih vrstah dreves na vsaki ploskvi posebej. Na severni raziskovalni ploskvi
sta to gaber in hrast, na južni pa jesen in hrast. Okrog vsakega debla je bil
speljan žlebič iz gumijaste polcevke. Na vsaki ploskvi posebej so se v desetih
posebnih košarah redno zbirale količine odpadlega listja za določitev indeksa
listne površine LAI. Z istim namenom in v istih točkah se je izvajalo
tudi hemisferično fotografiranje krošenj in tri serije meritev fotosintetskega
aktivnega sevanja PAR. Za potrebe določanja indeksa listne površine je
bila najpogostejšim vrstam dreves določena specifična površina listov
SLA.
Hkrati z vsemi temi meritvami so se v okviru celotnega projekta izvajale tudi
ostale meritve na povodju, ki so tudi uporabljene v nalogi.
Indeks listne površine
LAI se je določal po treh
metodah. Metodi hemisferičnega fotografiranja in merjenja fotosintetskega
aktivnega sevanja PAR v obdobju z listjem podcenjujeta vrednosti
LAI.
Za izračune so bili privzeti rezultati najnatančnejše metode zbiranja odpadlega
listja. Dobljeni vrednosti za LAI
po tej metodi sta 6,7 za južno pobočje
in 7,3 za severno. Manjša vrednost LAI na južnem pobočju je bila
pričakovana, saj ima gozd na tem pobočju bolj odprt značaj kot na severni
strani.
Analiza meritev je pokazala, da obravnavano obdobje po
količini padavin precej odstopa od 35-letnega povprečja, zlasti v jesenskih
mesecih, ko je padlo kar tretjina celotne letne vsote padavin. Intenziteta
padavin je bila največja poleti in najmanjša pozimi. Regresijske analize padavin
izmerjenih na posameznih dežemerih so pokazale dobro ujemanje s koeficientom
korelacije nad 0,95. Regresijske analize meritev deleža padavin, ki padejo skozi
odprtine med krošnjami ali kasneje z njih, s stalnimi in premičnimi merilci, so
pokazale pričakovane višje vrednosti premičnih merilcev z zelo visokimi
koeficienti korelacije. Povprečna vrednost tega deleža padavin znaša 67,1 % za
južno in 71,5 % za severno raziskovalno ploskev. Izmerjene vrednosti so
primerljive z rezultati podobnih raziskav oz. so za približno 10 % nižje.
Povprečni delež odtoka po deblu znaša 4,5 % na južni strani in 2,9 % na severni.
Razlika je posledica različne strukture in lastnosti gozda na obeh pobočjih. Na
osnovi meritev in bilančne enačbe je bilo ugotovljeno, da znašajo povprečne
letne izhlapele prestrežene padavine 28,4 % za južno in 25,4 % za severno ležeče
pobočje. Torej več kot četrtina padavin padlih nad gozdom izhlapi nazaj v
ozračje. Vrednosti so precej visoke v primerjavi z izhlapelimi prestreženimi
padavinami listopadnih gozdov, ki se ponavadi gibljejo med 15 in 25 %.
Podobne rezultate dobimo tudi z različnimi modeli
prestreženih padavin. Pri modeliranju so bili uporabljeni tako obstoječi,
kot na novo izdelani modeli. Modelirane vrednosti so se v splošnem dobro ujele z
meritvami. Odstopanja so bila v mejah standardne napake. Izjema je le model, pri
katerem je bila intenziteta izhlapevanja z mokrih krošenj izračunana po
Penman-Monteithovi enačbi. Tako izračunane vrednosti intenzitete izhlapevanja so
močno podcenjene, kar so potrdile že mnoge raziskave v mokrih priobalnih
pogojih. Iz tega lahko povzamemo, da Penman-Monteithova enačba kljub svoji
natančnosti in visoki zahtevi po podatkih ni primerna za izračun izhlapevanja
prestreženih padavin z mokrih krošenj gozda v danih klimatskih pogojih.
V disertaciji je dokazan velik vpliv južnih vetrov na
količine izhlapelih prestreženih padavin. Iz tega razloga je bil izdelan nov
model, ki mu je bil dodan vpliv toplih južnih vetrov. S tem smo podcenjevanje
izhlapelih prestreženih padavin po Penman-Monteithu odpravili.
V zaključnem delu
disertacije je izdelana še razlika v vodni bilanci glede na rabo kmetijskih
površin v letih 1948 in 1988. Vzrok za razliko v bilanci so spremembe v količini
prestreženih padavin na površinah poraščenih z gozdom in obdelovalnih površinah.
Za prestrežene padavine na kmetijskih površinah je privzeta koruza, ker so bili
v literaturi parametri zanjo podani najnatančneje in ker so prestrežene padavine
koruze v primerjavi z drugimi poljščinami zaradi velike pokrovnosti in indeksa
listne površine sorazmerno velike. Izračuni so pokazali, da se je količina
površinskega odtoka med letoma 1948 in 1988, zaradi spremembe v rabi površin,
zmanjšala za 4,3*106 m3, kar predstavlja 40,5 %
zmanjšanje.
Ključne besede:
prestrežene padavine, indeks listne površine, LAI, Gash, modeliranje, Dragonja,
terenske meritve, hidrologija, veter
UDK:
551.506+551.508+556.12/.13.08+556.51+630.11(282.249Dragonja)(043.3)
Abstract
The thesis is a part of an
extensive scientific-research project “Dragonja: Forest – Water – Soil – Climate
Interactions”, which has been carried out by Vrije University from Amsterdam and
University of Ljubljana since 1999. The Dragonja watershed was chosen as an
experimental watershed, since it is interesting because of the intensive natural
reforestation in the last decades, which has caused a decrease in minimal and
maximal flows. At the same time no noticeable precipitation and temperature
changes have been perceived. The main intention of the project is to figure out
the impact of reforestation on the water balance of the entire watershed.
In the frame of the thesis, precise measurements and analysis
of individual components of the forest hydrological cycle were made.
Contemporary measurement equipment was used and different rainfall interception
models were applied on the experimental watershed of Dragonja. For this reason,
two forest plots in the deciduous forest above confluence of the Dragonja and
Rokava rivers were selected at the end of the year 1999 in co-operation with
Vrije University from Amsterdam. One is located on the north-facing slope (1419
m2) and other on the south-facing slope (615 m2).
Measurements started in autumn 2000. The thesis describes the result of a
one-year period from October 2000 till September 2001. Precipitation above the
canopy, throughfall and stemflow were measured on both research plots. All
mentioned quantities were measured automatically with digital recording of
results every 10 minutes. Rainfall above the canopy was measured with a tipping
bucket rain gauge and with a totalisator (manual gauge) for control. Throughfall
was measured with two steel gutters in combination with ten manual gauges, which
were emptied and moved randomly every time. The spatial variability of
throughfall was captured in this manner. Stemflow was measured on two most
typical species in each plot. In the north plot oak and hornbeam trees were
selected, and ash and oak trees in the south plot. Each tree was fitted with a
rubber collar around the stem. Litter was collected regularly in 10 baskets in
each plot for LAI estimation. For the same purpose and at the same
points, hemispherical photographs of canopies were taken and three series of
measurements of photosynthetically active radiation (PAR) were made. The
specific leaf area SLA was estimated for the main species for need of estimating
the leaf area index. At the same time in the frame of the entire project, other
measurements, used in this thesis, were carried out in the watershed.
The leaf area index
LAI was estimated by three
methods. The method of hemispherical photography and photosynthetically active
radiation PAR measurements underestimated
LAI values in the
foliated period. Results of litterfall collection, the most accurate method,
were used for further calculations. The obtained values of LAI were 6.7 and 7.3
for the south and north slope, respectively. A smaller value of LAI on the south
slope was expected, because of the more open character of this forest.
Analysis of the measurements showed that the observed period
deviates considerably in the amount of precipitation compared to the 35-year
average, especially in the autumn months, when one third of the entire annual
sum of precipitation fell. The largest intensity was in the summer months and
the smallest in winter. Regression analyses of precipitation measured on
individual rain gauges showed good agreement with a correlation coefficient over
0,95. Regression analyses of the measured throughfall with fixed and roving
gauges gave expected higher values of roving gauges with very high correlation
coefficients. The average value of this proportion of precipitation amounts to
67.1 and 71.5 % for the south and north research plot, respectively. The
measured values are comparable with the results of similar researches, or else
they are about 10 % lower. The average stemflow fraction amounts to 4.5 and 2.9
% of the associated rainfall for the southern and northern plot, respectively.
The difference is a consequence of the different structure and characteristics
of forests on both slopes. On the basis of the measurements and water balance
equation, it was calculated that the average annual interception losses amount
to 28.4 and 25.4 % for the south and north slope, respectively. So more than one
fourth of the precipitation falling above the forest evaporates back to the
atmosphere. The values are rather high compared with the interception losses of
deciduous forests, which are usually between 15 and 25 %.
Similar results were also obtained with different models of
interception. For modelling, both the existing and the newly created models were
used. The modelled values were in general well fitted to the measured values.
Deviations were within the standard error. The only exception was the model,
which calculated evaporation rate from the wet canopy based on the
Penman-Monteith equation. Those values were strongly underestimated, as has
already been confirmed by many researches in wet maritime conditions. We can
summarise that the Penman-Monteith equation, despite its accuracy and very high
data demand, is not appropriate for the calculation of interception losses from
a wet canopy in the given climatic conditions.
The thesis has proved the large impact of south winds on
amounts of interception loss. For this reason a new model was created, which
takes into account the influence of warm south winds. Thus, underestimation of
interception losses based on the Penman-Monteith equation was eliminated.
In the final part of the thesis the difference in water balance based on land
use in the years 1948 and 1988 was made. Changes in the amount of interception
on forested and agricultural areas result in differences in the water balance.
Corn was adopted as a reference crop on agricultural areas, because its
parameters were represented in the literature most accurately and because the
interception of corn on account of the large canopy cover and leaf area index
are relatively large compared with other crops. Calculations showed that the
quantity of surface runoff between the years 1948 and 1988, because of changes
in land use, was reduced by 4.3*106 m3, which represents a
40.5 % reduction.
Key words:
interception
precipitation, leaf area index, LAI, Gash, modelling, Dragonja, field
measurements, hydrology, wind
UDC:
551.506+551.508+556.12/.13.08+556.51+630.11(282.249Dragonja)(043.3)
PETKOVŠEK Gregor :
Kvantifikacija in modeliranje
erozije tal z aplikacijo na povodju Dragonje
Mentor: izr. prof. dr. Matjaž Mikoš
Somentor: izr. prof. dr. Matjaž Četina
zagovor: september 2002
Povzetek
Naloga
obravnava procese erozije tal in odplavljanja zemljin z eksperimentalnega
povodja Dragonje v celoti ali pa enega njegovih delov. Na podlagi meritev v
zadnjih treh letih smo določili stopnje sproščanja s klifov, ki so
najpomembnejši točkovni vir sedimentov. Merili smo tudi odplavljanje sedimentov
in padavine, ter določili njihovo erozivnost. Z modelom na podlagi enačbe RUSLE,
ki smo ga izdelali sami, smo ocenili sproščanje zemljin zaradi vodne erozije s
površinskih virov, med katerimi so najpomembnejše kmetijske površine. Uporabili
smo tudi zgodovinske podatke o padavinah in rabi tal, s pomočjo katerih smo
ocenili trende erozivnosti padavin in sproščanja zemljin v povodju.
Meritve so vključevale meritve padavin na petih lokacijah,
meritve pretokov in vzorčevanje vode na povodju Rokave (največji pritok Dragonje
v zgornjem delu povodja), ter meritve erozije na klifih. Vzorcem vode je bila
določena vsebnost trdnih snovi in na ta način določen pretok lebdečih plavin.
Pokazalo se je, da je skupni letni odtok lebdečih plavin okrog 1500 ton, od tega
klifi prispevajo nekaj 100 m3 sedimentov v obliki drobnejših frakcij.
Premeščanje rinjenih plavin se vrši bržkone le enkrat letno; letna prodonosnost
je reda 100 m3. Primerjali smo več metod za oceno koeficienta
odplavljanja. Kot najprimernješa se je pokazala prirejena metoda USLE-M.
Koeficient odplavljanja v merskem prerezu znaša slabih 10 %.
Erozivnost padavin na območju povodja Dragonje znaša med 2500
(osrednji del povodja) in 3500 MJ/ha mm/h (povirje). Na določenih lokacijah je
letna erozivnost presegla 5000 MJ/ha mm/h. Pri tem igra pomembno vlogo
mikrolokacija. Erozivnost padavin je največja tik nad sotočjem Dragonje in
Rokave, to je tam, kjer se iz doline strmo vzpne sleme, na katerem leži vas
Labor. Tam je lokalno največja ne le količina padavin, temveč tudi njihova
intenziteta. V obdobju meritev 2000-2002 je največja polurna intenziteta na tej
lokaciji presegla 100 mm/h. Močni nalivi z veliko erozivnostjo se pojavljajo
poleti in jeseni, medtem ko v zimskih in pomladanskih mesecih prevladujejo manj
intenzivna dolgotrajnejša deževja.
V okviru naloge smo razvili tudi enačbo za napovedovanje
mesečne erozivnosti padavin na podlagi dnevnih padavin. Pokazalo se je, da je
najustreznejši parameter vsota kvadratov dnevnih padavin, pomnožena s
koeficientom, ki se mesečno spreminja in približno sledi srednji mesečni
temperaturi. Na ta način smo lahko izvrednotili erozivnost padavin na lokacijah
in v obdobjih, ko še niso bili na razpolago podatki o padavinah v 5-minutnih
intervalih. Ugotovili smo, da v zadnjih desetletjih letna erozivnost padavin
upada.
Model erozije tal smo zasnovali tako, da ga
lahko vključimo v obstoječi Šifrant prispevnih površin Republike Slovenije.
Model temelji na enačbi RUSLE in na erozijskih enotah, od katerih ima vsaka
enotno rabo tal in vsa pripada isti prispevni površini. Model smo aplicirali na
zgornjem delu povodja. Pokazalo se je, da lahko največje sproščanje pričakujemo
na parcelah, kjer so vinogradi (50 t/ha/leto) ter sadovnjaki in njive (okrog 20
t/ha/leto). Primerjava z Gavrilovićevo enačbo je pokazala, da se vrednosti
razlikujejo za približno faktor 2. Če namesto mreže erozijskih enot uporabimo
mrežo pravokotnih celic (raster), so pri velikosti obravnavanega območja nad 10
km2 razlike zanemarljive. Glede na spremembe rabe tal in erozivnosti
padavin od šestdesetih let do danes smo ugotovili, da se je sproščanje zaradi
samo površinske vodne erozije tal prepolovilo, če pa upoštevamo še erozijo s
klifov, je zmanjšanje še večje (velja za podpovodje Rokave).
Ključne besede:
erozija tal,
Dragonja, eksperimentalno povodje, erozivnost padavin, modeliranje, RUSLE, GIS,
raba tal
UDK:
551.311.2:556.51(282.249 Dragonja)(043.3)
Abstract
The
thesis presents a study of the processes of soil erosion and sediment transport
in the Dragonja experimental catchment or some of its parts. Measurements in the
last three years were a basis for estimation of the sediment production from
cliffs, which are the most important point sources of sediment in the area.
Sediment yield and rainfall were also measured. A numerical model, based on
Revised Universal Soil Loss Equation, was developed to assess sediment
production from areal sources, especially agricultural areas. Historical data on
precipitation and land use was used to determine the trends in rainfall
erosivity and soil erosion rates in the catchment.
Measurements included rainfall measurements on five
locations, discharge measurements and sediment sampling in the Rokava catchment
(the biggest tributary in the upper part of the Dragonja catchment) and
monitoring of sediment production from cliffs. The discharge of suspended
sediment was determined from suspended sediment samples and turned out to be
around 1500 tonnes per year. Cliffs contribute a few 100 m3 of
suspended sediment per year to the stream. Annual bed load transport can be
estimated to about 100 m3 of bed load sediment per year. Sediment
yield and delivery ratio was calculated using different procedures. A variation
of USLE-M gave best results. Sediment delivery ratio in the gauged cross section
was found to be around 10%.
Rainfall erosivity in the Dragonja catchment varies between
2500 MJ/ha mm/h (central part) and 3500 MJ/ha mm/h (upper part). In the year
2001, erosivity exceeded 5000 MJ/ha mm/h in one of the monitored locations just
above the confluence of Dragonja and Rokava. At this point, a ridge rises
steeply from the valley. There, not only the highest amount of precipitation has
been recorded, but also the highest intensity. In the monitored period from
autumn 2000 till now, the highest half-hour intensity in this location exceeded
100 mm/h. Intensive storms with high rainfall erosivity occur during summer and
autumn, while winter and spring storms tend to be longer and have lower
intensity.
Furthermore, the equation for calculating monthly rainfall
erosivity form daily precipitation data was developed. The most suitable
parameter was found to be the sum of squares of daily precipitation, multiplied
by a coefficient that varies from month to month, and approximately follows the
mean temperature of a month. This equation enabled us to determine rainfall
erosivity in the periods and the locations where 5-minute interval precipitation
data was not available. This way, long term trends in rainfall erosivity were
determined, which show a slow decline during the last few decades.
Soil erosion model was developed in such a way that it can be used with the
existing Watershed Coding System of the Republic of Slovenia. The base of the
model is the RUSLE equation, applied to erosion units. Each erosion unit has a
uniform land use. The model was applied to the upper part of the catchment. The
results show that the highest soil loss can be expected in vineyards (50
t/ha/y), orchards and fields (20 t/ha/y). A comparison with the Gavrilović
equation showed that the latter gives the rates that are about half of those of
RUSLE. If square grid modelling approach is applied instead of erosion units,
the differences are very small when models are applied to an area of the size of
10 km2. The change in erosion rate from the sixties till present was
determined based on land use and rainfall erosivity trends. Erosion rate form
areal sources has halved, while erosion rate from cliffs (refering to Rokava
subcatchment only) has decreased even more.
Key words:
soil erosion, Dragonja,
experimental catchment, rainfall erosivity, modelling, RUSLE, GIS, land use
UDC:
551.311.2:556.51(282.249 Dragonja)(043.3)