Intercomparison of PWV from MODIS, radiosonde and GPS at Ankara TUSAGA-Active station


DENİZ İ., GÜRBÜZ G., MEKİK Ç.

IGS Workshop 2017, Paris, France, 3 - 07 July 2017

  • Publication Type: Conference Paper / Summary Text
  • City: Paris
  • Country: France
  • Hacettepe University Affiliated: No

Abstract

Water vapor is essential due to its role in determining the distribution of clouds. Therefore, to investigate the climate, meteorologists and climatologists monitor the distribution of the water vapor in the atmosphere. There are several observation systems to estimate precipitable water vapor (PWV): radiosonde, GPS and Moderate Resolution Imaging Spectroradiometer (MODIS). Radiosondes monitor the water vapor with a high accuracy and vertical resolution, but with low temporal and spatial resolution. On the other hand water vapor is estimated using GPS at a high level of accuracy with high temporal and spatial resolution. It is based on the computation of the tropospheric zenith delay (ZTD) which consists of wet (ZWD) and dry (ZHD) components. The first component (ZWD=ZTD-ZHD) is converted to the water vapor by the weighted mean temperature equation (Tm) or the conversion factor (Q). The last one MODIS on NASA Terra and Aqua platforms collect global water vapor daily with a spatial resolution of 1x1 km (at nadir) and an accuracy of 5-10%.

In this study, PWV obtained from MODIS of Ankara TUSAGA Active station (ANRK) (PWVM) is compared with the co-located radiosonde station (PWVR) and GPS estimated PWV values (PWVG) from 1 June 2013 to 31 December 2013. Here, PWVG is estimated with Tm (PWVGT) and QBEU (PWVGQ) models applicable for Turkey. Aside from PWVG and PWVR, PWVM is obtained from the Terra MODIS near-infrared water vapor products (MOD05_L2) under cloud free conditions and cloudy conditions. Also, the PWVM acquisition times are not congruent with those of GPS and radiosonde, so the cubic splines interpolation is applied to PWVGT, PWVGQ and PWVR values to obtain those at MODIS acquisition time. The differences of PWVM under the cloudy conditions relative to PWVR are smaller than those between PWVM and PWVGT, PWVM and PWVGQ. The results show that PWVM and PWVR (91%) agree better with each other than with PWVGT, PWVGQ (~87%). It is also shown that PWVR is in good agreement with PWVGT, PWVGQ (%87). Moreover, the differences of PWVM under the cloud free conditions relative to PWVR are also smaller than those between PWVM and PWVGT, PWVM and PWVGQ. The results indicate that PWVM is compatible with PWVR, PWVGT, PWVGQ (93%). It is also revealed that PWVR agrees with PWVGT, PWVGQ (%90).