Detections and simulations of tropospheric water vapor fluctuations due to trapped lee waves by ALOS-2/PALSAR-2 ScanSAR interferometry

Detailed wave-like spatial patterns of atmospheric propagation delay signals associated with mountain lee waves were detected in Hokkaido and Tohoku by synthetic aperture radar (SAR) interferometry (InSAR) with the ScanSAR mode observation data of a Phased Array-type L-band Synthetic Aperture Radar 2 on board the Advanced Land Observing Satellite 2. Both cases occurred under stable atmosphere conditions. The InSAR-observed peak-to-trough line of sight changes in the mountain wave signals was 4 and 5 cm with the horizontal wavelengths of 9 and 15 km in Hokkaido and Tohoku, respectively. Locations of positive phase maxima in the mountain wave signals coincides with locations of cloud streets observed by visible satellite imagery, indicating that crests of mountain waves contain relatively much water vapor compared with wave troughs. Numerical weather simulations with the horizontal grid spacing of 1 km were performed to reproduce InSAR phase variations, and as a result those simulations could reasonably reproduce observed wave amplitudes and wavelengths in both cases. On the other hand, numerical simulations tended to overestimate wave attenuation rates: simulated mountain waves decreased as the wave propagated faster than those of observed signals. Because the simulated wave attenuation rate is sensitive to physics in the planetary boundary layer (PBL), we investigated the reproducibility of five PBL schemes implemented in the WRF model. As a result, all the PBL schemes showed little attenuation except for the Yonsei University scheme (YSU), while the wavelength in the YSU was most close to the observation. Our study demonstrated the uniqueness and usefulness of InSAR for meteorological application as the ability to map the detailed water vapor distribution regardless of cloud cover. In addition, the reasonable reproducibility of the water vapor delay signal due to lee waves by the numerical weather model encourages researchers who tackle the correction of the tropospheric propagation delay, increasing the accuracy in detecting surface deformations.