11:15 AM - 11:30 AM
[STT36-09] Error level evaluation of L-Band InSAR Water Vapor Observation by Comparison with GNSS Observations
Keywords:InSAR, ALOS2, water vapor, error propagation
Interferometric synthetic aperture radar (InSAR) can measure not only the surface deformation, but also the information of troposphere at the high spatial resolution This is caused by the propagation delay effect similar to the GNSS (Global Navigation Satellite System), which is approximately proportional to the zenith total delay (ZTD) or the precipitable water vapor (PWV). Previous researches have evaluated the precision of InSAR PWV observation using C-band SAR by comparing with GNSS (Tang et al. 2016, Mateus et al. 2020). In general, each satellite SAR sensor has a single frequency band. Most of all SAR satellites mainly use three types of frequency bands such as X-band, C-band, and L-band. The precision of meteorological observation data is necessary to calculate more accurate initial condition when observations are assimilated in weather forecasting. It is necessary to evaluate the precision of water vapor observation using InSAR data in each frequency band because each band has different characteristics in coherence and sensitivity to the ionosphere.
In this work, we evaluated the error level of water vapor observation by L-band InSAR. At first, we calculated the distribution of the zenith total delay difference (ΔZTD) from the InSAR data by the projection from the slant path delay using the simple trigonometric function. The overall offset of InSAR ΔZTD was corrected by the ΔZTD obtained from GNSS observations in the scene. Then, we computed the standard deviation of the residuals between the InSAR ΔZTD and the GNSS ΔZTD. The ΔZTD was then converted to the ΔPWV with a constant coefficient. Finally, we calculated the standard deviation of InSAR-derived PWV at a single epoch using the law of error propagation. Here we assumed that the noise level of water vapor observations on each SAR acquisition was at the same level, and the absolute error level of the GNSS ZTD was set to 16.1 mm that was obtained by a previous research (Boccolari et al. 2002). We used the GNSS Earth Observation Network System (GEONET) data processed with the 5-minutes PPP analysis by the Nevada Geodetic Laboratory at University of Nevada, Reno.
In the InSAR processing, we adopted the Split-Spectrum Method for compensating the differential ionospheric path delay in L-band interferograms. SAR data used were observed by the ALOS-2/PALSAR-2 stripmap mode. The data were collected over four areas in Japan: southern Ibaraki, western Tokyo and Kanagawa, Osaka, and southern Kyushu, spanning from 2014 to 2020. The multilook processing resulted in reduction of the decorrelation noise with sacrificing the spatial resolution to 100×100 m.
Our results showed that the overall distribution of residuals was unbiased and the standard deviation were 8.64 mm in ΔZTD. Estimated standard deviations from the residuals were different in each region, and tended to be larger for the data observed during the rainy season (from June to September). We evaluated the precision of the PWV measurement by L-band InSAR as 2.83 mm, indicating the precision of the L-band InSAR PWV observation is as good as other PWV measurements like C-band InSAR, GNSS, radiometer and weather satellites. By the time of the presentation, we will perform the time series analysis to evaluate the precision of the InSAR PWV measurement for each SAR acquisition days.
In this work, we evaluated the error level of water vapor observation by L-band InSAR. At first, we calculated the distribution of the zenith total delay difference (ΔZTD) from the InSAR data by the projection from the slant path delay using the simple trigonometric function. The overall offset of InSAR ΔZTD was corrected by the ΔZTD obtained from GNSS observations in the scene. Then, we computed the standard deviation of the residuals between the InSAR ΔZTD and the GNSS ΔZTD. The ΔZTD was then converted to the ΔPWV with a constant coefficient. Finally, we calculated the standard deviation of InSAR-derived PWV at a single epoch using the law of error propagation. Here we assumed that the noise level of water vapor observations on each SAR acquisition was at the same level, and the absolute error level of the GNSS ZTD was set to 16.1 mm that was obtained by a previous research (Boccolari et al. 2002). We used the GNSS Earth Observation Network System (GEONET) data processed with the 5-minutes PPP analysis by the Nevada Geodetic Laboratory at University of Nevada, Reno.
In the InSAR processing, we adopted the Split-Spectrum Method for compensating the differential ionospheric path delay in L-band interferograms. SAR data used were observed by the ALOS-2/PALSAR-2 stripmap mode. The data were collected over four areas in Japan: southern Ibaraki, western Tokyo and Kanagawa, Osaka, and southern Kyushu, spanning from 2014 to 2020. The multilook processing resulted in reduction of the decorrelation noise with sacrificing the spatial resolution to 100×100 m.
Our results showed that the overall distribution of residuals was unbiased and the standard deviation were 8.64 mm in ΔZTD. Estimated standard deviations from the residuals were different in each region, and tended to be larger for the data observed during the rainy season (from June to September). We evaluated the precision of the PWV measurement by L-band InSAR as 2.83 mm, indicating the precision of the L-band InSAR PWV observation is as good as other PWV measurements like C-band InSAR, GNSS, radiometer and weather satellites. By the time of the presentation, we will perform the time series analysis to evaluate the precision of the InSAR PWV measurement for each SAR acquisition days.