| Summary: | The aim of this study was to detect vertical crustal movements in Estonia and find out possible changes of vertical crustal movements in time. Vertical velocities of the benchmarks were calculated from the precise levellings between 1933 and 2011 of the Estonian levelling network using the joint weighted kinematic least squares adjustment of the levelling campaigns. Two different mathematical models, the “heights included” and the “heights eliminated” model, were used in the adjustment. Different options of the computer software Surfer were used for modelling of the vertical crustal movements. Accuracy of the models was estimated by finding differences between the velocities interpolated from the models and adjusted vertical velocities of the benchmarks, using the cross validation technique, and by comparing models with the results from other geodetic measurements (continuously operating GNSS stations, tide gauges, other land uplift models). From the variance component estimation, it appeared that levelling errors of the First levelling campaign were ~3 times larger than estimated a priori. Final adjustment was performed with the re-scaled weights according to the results of the variance component estimation. Models of the vertical crustal movements EST2013LU and EST2015LU were created based on the vertical velocities of the benchmarks. According to the models, rates of the land uplift in Estonia range from −0.7 mm/yr in SE Estonia to +2.8 mm/yr in the island of Hiiumaa in NW Estonia. Accuracy of the models was estimated to be ±0.4 mm/yr on average. The comparison of the models with the velocities from the independent measurement methods revealed best fit with the velocities of the GNSS permanent stations where residual differences remained within ±0.3 mm/yr on average. The discrepancies between the velocities of the coastal tide gauges and the velocities from the models were ±0.7…±1.0 mm/yr on average. Obtained differences implied to the systematic biases in tide gauge velocities. Comparison with the historical vertical crustal movement maps of Estonia showed that differences remained within ±0.7 mm/yr on average. The fit between the most recent Fennoscandian LU map NKG2005LU and the models obtained in the present study was very good. Differences remained within ±0.3 mm/yr on average. It appeared also that vertical velocity of the benchmarks has been significantly changed between the levelling periods. Results of the study can be used to estimate the risks to the coastal areas coming from the global warming related rise of the sea level. Doktoritöö eesmärk oli leida maakoore vertikaalliikumiste kiirused Eestis nelja kordusnivelleerimise (1933-2011) andmete põhjal ja välja selgitada reeperite kiiruste muutumine ajas. Reeperite vertikaalliikumise kiirused leiti nivelleerimiste ühisest kaalutud kinemaatilisest tasandusest vähimruutude meetodil. Tasandusel kasutati kahte matemaatilist mudelit: nn „kõrgustega“ ja „kõrgusteta“ mudelit. Maakoore vertikaalliikumiste modelleerimiseks kasutati tarkvara Surfer erinevaid võimalusi. Mudelite täpsust hinnati mudelist interpoleeritud ja reeperite tasandatud kiiruste vaheliste erinevuste leidmise, ristvalideerimise ja sõltumatute mõõtmistulemustega (GNSS-püsijaamad, veemõõdujaamad, teised mudelid) võrdlemise teel. Dispersioonikomponentide hindamisest selgus, et esimese nivelleerimiskampaania vead on ~3 korda suuremad kui a priori eeldati. Lõplik tasandus teostati dispersioonikomponentide hindamise tulemuste põhjal ümberskaleeritud kaaludega. Reeperite kiiruste põhjal loodi Eesti maakoore vertikaalliikumiste mudelid EST2013LU ja EST2015LU. Mudelite põhjal ulatuvad maatõusu kiirused Eestis alates −0.7 mm/a Kagu-Eestis kuni +2.8 mm/a Hiiumaal. Mudelite täpsuseks hinnati keskmiselt ±0.4 mm/a. Võrdluses sõltumatutest meetoditest määratud maatõusu kiirustega selgus, et parim oli sobivus GNSS-püsijaamade kiirustega, keskmiselt ±0.3 mm/a. Halvim oli sobivus ranniku veemõõdujaamade kiirustega: ±0.7….±1.0 mm/a. Saadud erinevused viitasid süstemaatilistele nihetele veemõõdujaamade kiirustes. Võrdlus Eesti varasemate maakoore vertikaalliikumiste kaartidega näitas, et erinevused jäid keskmiselt ±0.7 mm/a piiridesse. Mudelite sobivus viimase Fennoskandia maatõusu mudeliga NKG2005LU oli aga väga hea, erinevused jäid keskmiselt ±0.3 mm/a piiridesse. Samuti selgus, et nivelleerimisperioodide vahel on reeperite kiirus statistiliselt oluliselt muutunud. Uurimistöö tulemusi saab kasutada kliimasoojenemisest tulenevate meretõusu riskide hindamiseks rannikualadele
|
|---|