Establishing Zenith Wet Delay model (ZWD) and developing a framework for generating high resolution PWV for extreme weather monitoring using MT-InSAR and GNSS for Indian Himalayan region

Precipitable water vapor (PWV), an essential climate variable enlisted by the Global Climate Observing System (GCOS), can be efficiently mapped using popular Earth Observation techniques like Global Navigation Satellite System (GNSS) and Interferometric Synthetic Aperture Radar (InSAR). When used in a combination, these techniques complement each other in terms of temporal and spatial resolution for producing high-resolution PWV. Tropospheric water vapor (measured as wet delay) can be perceived in two components: non-turbulent (ZWD_NT) as well as turbulent information (ZWD_T). Wet delays from InSAR acquisitions are able to capture only the turbulent information, hence need to be enhanced. The non-turbulent component further can be spatially classified into shortwave and longwave components. In this study, GNSS observations from a dense network of twelve GNSS CORS for 2021, from the newly established CORS network by Survey of India with a homogeneous spread over Uttarakhand (UK) state, is used to establish the ZWD_NT model. We develop an exponential elevation-dependent model for shortwave components, incorporating seasonal variations and a location-dependent model for long wave components. Model assessment shows the performance of the developed model when a satisfactory mean RMSE of 8.32 mm is obtained through internal checks, which shows the efficacy of the developed model in capturing elevation dependency and seasonal variations. Further, a geodetic framework is conceptualized wherein the values derived from developed ZWD_NT model are appended to non-differential ZWD_T estimated after Small BAseline Subset InSAR (SBAS-InSAR) processing at measurements points density of about 50 million points from 30 ascending pass Sentinel 1A acquisitions, to arrive at full atmospheric information (ZWD_total). A previously developed weighted mean temperature (Tm) model for the highly undulating himalayan foothills region in the UK, is incorporated in the conversion of ZWD_total to PWV, for better accuracy and further assessment. A high resolution combination of PWV derived from complementary GNSS and InSAR techniques can be efficiently utilized in improving the numerical weather prediction (NWP) skill as well as monitoring extreme weather event since the spatial variations in local tropospheric conditions of a hilly terrain are quite frequent. When validated against PWV from ERA5 reanalysis data, a mean RMSE of 9.5 mm is obtained, except for the monsoon period, when RMSE falls in the range of 10-20mm. This may be due to the fact that InSAR partial non differential PWV captures the spatially correlated artifacts especially in the temporal vicinity of a rainy event. The results show that the proposed approach can effectively enhance the InSAR derived non-differential PWV and provide useful information at a high spatial resolution in a varied topography in lesser Himalayan.These high-resolution PWV maps hold great promise for enhancing meteorological understanding and quantitative analysis, specially during heavy rainfall in a complex terrain like UK. With the upcoming NISAR mission for the Indian subcontinent, spatio-temporal analysis of tropospheric parameters can be further enhanced for weather forecasting.