Back to search

National and Kapodistrian University of Athens

Επίπτωση της κλιματικής αλλαγής στις συνιστώσες του υδρολογικού κύκλου

Abstract

dc:description

This PhD research is placed in the large area of studying the influences of climate change on the components of the hydrological cycle and especially in evaporation (or latent heat flux). The main purpose was the estimation of evapotranspiration using not only simple conventional ground measurements, but simple models based on remote sensing data along with the climate change and the decreasing/increasing trend of solar radiation (Global Dimming and Brightening phenomenon, GDB). In order to contribute to the theoretical concern on the consistency in evaporation trends and trends of solar radiation or humidity and wind speed, reference evapotranspiration (ET) was estimated from REMCT (Reference Evapotranspiration Model for Complex Terrains) model. REMCT model was applied in Greece after developing a methodology for calibrating the model parameters for various groups of stations and it was independently validated from available pan measurements. GDB phenomenon was detected in Greece and especially in the urban area of Athens. Furthermore the consistency in ET trends and the decreasing/ increasing trends of solar radiation and the possible causes of GDB phenomenon (such as, changes in aerosols concentration or cloudiness or humidity at the atmosphere) were investigated. Additionally, the impact of GDB phenomenon on pan evaporation changes as compared to the effects of temporal changes of other meteorological parameters was examined. After verifying the ability of the calibrated REMCT model to incorporate the effects of GDB phenomenon, its use was investigated, for projecting future trends in ET using projections of regional climate model and for estimating ET from MODIS Land Surface Temperature (LST) data in Greece.Analytically, an ET calculation algorithm was developed from monthly precipitation and temperature measurements using REMCT model, which also includes the effect of climate and site elevation. For improving ET estimates from REMCT model, a methodology for calibrating the model parameters was developed using monthly estimates of potential evapotranspiration by Penman-Monteith method (ETPM) for 29 stations in Greece. The locally calibrated ET estimates (ΕΤDB) have been validated from pan evaporation (Epan) measurements and REMCT model was considered as suitable to estimate accurately the monthly ET and hence to investigate its time evolution for longer time periods. ΕΤDB trends in Greece were compared with sunshine duration (SD), trends “used as a proxy” of surface solar radiation, in Athens. The time evolution of ΕΤDB was explained by the GDB phenomenon, i.e. by its agreement with the decrease and the subsequent increase trend of solar radiation. REMCT model was also successfully applied in the station of National Observatory of Athens, using regional calibrated parameters and ΕΤDB estimates were verified from available monthly ETPM estimates. There has been a special investigation of the GDB phenomenon in the urban area of Athens highlighting the dimming/brightening sub-periods on annual and seasonal basis. Furthermore, the effect of cloudiness was linearly removed from SD measurements (“used as a proxy” of surface solar radiation) and causes of GDB phenomenon linked (mostly) to aerosols changes were detected. The ΕΤDB trends were examined during the period 1951-2001 and they were found in agreement with the detected dimming/brightening periods.GDB phenomenon and its impact on pan evaporation (Epan) were investigated widely in Greece. Causes of GDB were also examined by focusing especially on changes in aerosols, cloudiness (CL) and precipitable water (PW). For this purpose, daily PW data, from the NCEP-1 reanalysis data were extracted for 28 stations during the period 1974-2004 in order to examine their contribution in GDB phenomenon. Additionally, the consistency in SD anomalies or residuals SD time series and PW and CL anomalies, were investigated according to the dimming/brightening periods. Moreover, the implications of GDB phenomenon in Epan changes were also examined on yearly and seasonal basis for 13 stations during the period 1979-2004, where pan evaporation measurements were available. Epan anomalies and anomalies of meteorological parameters, such as SD, wind speed and vapor pressure deficit, were calculated in order to investigate the causes of Epan changes. The results indicated the SD changes as the most likely causative meteorological factor affecting Epan in Greece. Since REMCT model could incorporate GDB phenomenon, the use of ETDB estimates as an ET formulation based on minimum data, was investigated in order to predict ET changes in a future climate. Epan and ETDB (or ETPM) have been found well correlated with similarities between the magnitudes and signs of their trends (in various time scales). Daily projections of temperature and precipitation from the Regional Atmospheric Climate MOdel (RACMO2) were employed in REMCT model to estimate reference evapotranspiration (ETDB proj) in 29 stations in Greece during 1979-2004 (control period). The DBS (Distribution Based Scaling) method is used for the bias corrections of temperature and precipitation projections and the corrected values are used in the estimation of ETDB (ETDB cor). Furthermore, the corrected projections of the RACMO2 model were employed in the calibrated REMCT model and ETDB cor trends were also estimated for the future period (2011-2100) and compared with the corresponding ones of the control period. Positive trends were found for ETDB cor, accompanied mostly with negative precipitation trends and increasing temperatures during future period.Daytime and nighttime MODIS land surface temperature data (LST) were tested as an alternative input for maximum (Tmax) and minimum (Tmin) air temperature in three reference evapotranspiration (ET) models in Greece. Thus, daily Aqua MODIS LST data were extracted during the day (LSTday) or during the night (LSTnight), for 28 sites in Greece during the growing period (May- October) of the years 2002, 2003 and 2004. Firstly, the use of LSTday or LSTnight as “indicators” of ground Tmax or Tmin air temperatures, respectively, was investigated and a good correlation was found between MODIS LSTnight and Tmin. Then MODIS LSTday and LSTnight data were employed in the air temperature range-based Hargreaves-Samani method and in two other ET methods, depended upon precipitation besides air temperatures, the Droogers-Allen method and the calibrated REMCT model. The obtained ET results for 28 stations (using MODIS LST data) were compared with the corresponding ET estimates from the ground air temperature data and furthermore with Penman Monteith ET estimates based on ground data and as a result, REMCT model has shown the best ET estimates. In order to furthermore improve the accuracy of Tmax (as obtained from the LST data), the temperature-vegetation index (TVX) method was applied in various stations in Greece and it was evaluated at different time periods and areas. Thus, daily MODIS data of surface reflectance were used for the calculation of daily Normalized Difference Vegetation Index (NDVI). Maximum air temperature was estimated under the assumption that the temperature of a full cover canopy approaches the temperature of air within the canopy and therefore the normalized difference index of a full canopy (NDVImax) was estimated in Greece. Moreover, monthly maximum temperature estimates from the TVX method (TVX LSTday) were validated on temporal basis and the results have shown that the TVX method has highly improved the accuracy of monthly maximum temperature estimates from MODIS LSTday data. The method has also been validated on spatial basis using the estimated value of NDVImax in various stations in Greece improving slightly the accuracy of Tmax values. When the TVX LSTday estimates were used in ET models, a slight improvement was also observed compared either with ET estimates based on simple MODIS products or with ETPM estimates. In areas with lack of ground temperature data, MODIS products can be employed especially in REMCT model for estimating ET, achieving spatial analysis of ET at higher spatial resolutions.

Degree

thesis:*
Grantor dc:publisher
National and Kapodistrian University of Athens
Year dc:date
2015

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Kitsara, Giannoula
  • Κιτσαρά, Γιαννούλα

Subjects

dc:subject × 24

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/36724

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Kitsara, Giannoula; Κιτσαρά, Γιαννούλα. Επίπτωση της κλιματικής αλλαγής στις συνιστώσες του υδρολογικού κύκλου. National and Kapodistrian University of Athens, 2015. http://hdl.handle.net/10442/hedi/36724