Global atmospheric water vapour determination from GNSS and its applications for climate studies

Wang, X 2017, Global atmospheric water vapour determination from GNSS and its applications for climate studies, Doctor of Philosophy (PhD), Mathematical and Geospatial Sciences, RMIT University.

Document type: Thesis
Collection: Theses

Title Global atmospheric water vapour determination from GNSS and its applications for climate studies
Author(s) Wang, X
Year 2017
Abstract Water vapour is a principal greenhouse gas component in the atmosphere and its variation in concentration strongly influences the change of climate and weather. The long-term trend in water vapour has been regarded as a very important resource of atmospheric information for climate studies. However, precisely quantifying the distribution and variation of water vapour at a high spatiotemporal resolution is often a challenge if water vapour is retrieved only from traditional meteorological sensors due to their low spatiotemporal resolutions. Nowadays, using Global Navigation Satellite Systems (GNSS) to remotely sense precipitable water vapour (PWV) contents in the atmosphere has heralded a new era for weather and climate research. From GNSS observables the tropospheric zenith total delay (ZTD) of the GNSS signal can be estimated, which can be converted into PWV with meteorological measurements (surface pressure, temperature and humidity profiles).

Since most GNSS stations were built in or after 1990s, GNSS-derived PWV time series are not sufficiently long for investigation of some long-term climate change (e.g. global warming). However, those PWV time series that are longer than a 10-year period can be used to study some of short-term climate phenomena (e.g. EI Niño–Southern Oscillation, ENSO). Studies indicated that the frequency of ENSO occurrence would be doubled in future in response to global warming, which will have profound socio-economic consequences. However, a better understanding of the evolution of ENSO and accurately predicting its occurrence and consequences is still a big challenge to climate scientists. Therefore, obtaining and investigating additional observations for different atmospheric and oceanic variables are of significance in helping us better understand the evolution of this phenomenon and more accurately predict its consequences (e.g. floods and droughts).

In this study, we investigated the potential of using GNSS-derived PWV to study some climate patterns (e.g. ENSO) and to predict the consequences such as floods and droughts. A new long-term of PWV dataset was developed using GNSS-derived ZTDs provided by the Centre for Orbit Determination in Europe (CODE) and meteorological data from ERA-Interim (European Centre for Medium-Range Weather Forecasts Re-Analysis), which is an accurate global atmospheric reanalysis. Results showed a 0.5 % relative error and a 2.1 mm RMS error in the obtained water-vapour-weighted mean temperature and zenith hydrostatic delay (ZHD) respectively. The comparison between monthly PWVs derived from GNSS and co-located radiosondes (as the reference) at 12 stations showed a −0.23 mm bias and a 0.66 mm RMS error. This accuracy is sufficient for climate studies.

In this study, a time series analysis method called singular spectrum analysis (SSA) was adopted to analyse non-linear trends in the PWV time series obtained. SSA is a data-driven technique that uses time domain data to extract information from noisy time series without the use of a priori knowledge of physical phenomena in the time series. A new interpolation method named SSA for missing data (SSAM) is proposed to interpolate missing data and a gross error detection method called SSA with interquartile range (IQR) is proposed to detect gross errors in the time series. The complete and clean time series are then analysed using the SSA method. The non-linear trends in the PWVs estimated at many of the selected 56 GNSS stations located close to the sea indicate that they are significantly affected by ENSO events. Since ENSO events are monitored through observing the variation of sea surface temperature (SST), a comparison between PWV and SST indicates that generally, a 1 K increase in SST will lead to a 11% increase in PWV across all these stations. A case study at the TOW2 station (in Australia) shows that the variation in the PWVs has a very good agreement with total precipitation, and the non-linear trend in the PWVs clearly depicts the evolution of two severe floods and one severe drought in the region. These results suggest that GNSS-derived PWV together with other climatic variables (e.g. SST) can be used as a tracer of the evolution of ENSO events and a possible indicator of droughts and floods.

Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Mathematical and Geospatial Sciences
Subjects Climatology (excl. Climate Change Processes)
Navigation and Position Fixing
Tropospheric and Stratospheric Physics
Keyword(s) GNSS
Water Vapor
Climate Change
Time Series Analysis
Version Filter Type
Access Statistics: 276 Abstract Views  -  Detailed Statistics
Created: Fri, 21 Jul 2017, 10:51:44 EST by Denise Paciocco
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us