Precipitable water vapor characterization in the coastal regions of China based on ground-based GPS

Abstract

Water vapor plays an important role in climate change; thus, studying the spatial distribution and temporal variation of precipitable water vapor (PWV) in the coastal regions of China would help researchers to understand the climate characteristics of those regions. In this paper, 6-year 1-h interval PWV were derived from 27 Global Positioning System stations observations of Chinese coastal GPS observation network, surface meteorological data and European Center for Medium-Range Weather Forecasts (ERA-Interim) reanalysis products. The present study provides the use of these data to investigate the spatial-temporal variability of water vapor throughout the coastal regions of China. Latitude is the main factor affecting the spatial distribution of GPS-derived PWV; that is, PWV decreased by about 1.5mm for each 1° increase of latitude. For regions at the same latitude, a region that is relatively close to the ocean will have a higher content of PWV. The PWV in the southeastern and southwestern coastal regions of China is significantly higher in summer; this may be influenced by the southeastern and southwestern water vapor inflow corridors. The PWV obviously varies monthly, reaching a minimum in January; however, the timing of the maximum varied but usually appeared in June, July or August and was affected by the monsoons. The PWV varies largely between summer and winter with a larger gradient of change in PWV with latitude in winter than in summer. The positive correlation coefficient between PWV and the surface temperature varied in different seasons; this is related to the changes of temperature and the horizontal motion of water vapor. Use of the Fast Fourier Transform method showed that the PWV time series data have multi-scale characteristics. The amplitude and phase of the PWV time series in annual, semiannual, four month and seasonal cycles were extracted through harmonic wave analysis. The amplitude of four month and seasonal cycles did not pass significance testing. The maximum amplitude of the annual PWV cycle is located in mid-latitudes with 22.09mm, which gradually decreases towards high latitude and equator area. The peak time of annual PWV cycle occurs in July, which does not agree with the timing of the maximum monthly average PWV. The semiannual cycle amplitudes is between 0.42 and 6.32mm, with no significant correlation between their distribution and latitude. The peak time of semiannual PWV cycles is during about January–March and July–September, and the peak time gradually becomes late from north to south.