Modelling the height dependence of the wet path delay using ERA5 model-level fields

Telmo Vieira (University of Porto, Faculty of Sciences, Portugal)


M. Joana Fernandes (University of Porto, Faculty of Sciences, Portugal); Clara Lázaro (University of Porto, Faculty of Sciences, Portugal)

Event: 2018 Ocean Surface Topography Science Team Meeting

Session: Instrument Processing: Propagation, Wind Speed and Sea State Bias

Presentation type: Type Poster

Accurate determination of sea surface height (SSH) from satellite radar altimetry depends on the accuracy of all terms involved in its computation, namely the wet tropospheric correction (WTC). With a high spatial-temporal variation, this correction is best determined from on-board Microwave Radiometers (MWR) measurements, which become invalid close to land surfaces. Originally designed for open-ocean, satellite altimetry has been used over coastal and inland water zones, however measurements derived from MWR cannot be used over these regions. Alternative WTC sources over these regions are the Global Navigation Satellite Systems (GNSS) and the global atmospheric models, e.g. from European Centre for Medium-Range Weather Forecasts (ECMWF).
Designed for studies over the ocean, altimetric missions are mainly focused on the sea surface. For this reason, MWR-derived WTC measurements refer to the sea level, while WTC derived from an atmospheric model are computed at the level of its orography, usually a smoothed representation of a digital elevation model (DEM). The corresponding corrections from GNSS are computed at station height. Due to the differences between these data types, the modelling of the height dependence of the WTC is a crucial step to combine these different data (MWR, atmospheric model and GNSS) for altimetry studies over coastal and continental water regions. Over coastal zones, all measurements must refer to the sea level, while over continental waters they must refer to the level of the corresponding water body.
Therefore, an expression to reduce the WTC from GNSS station height and orography level to sea level (over coastal zones) and to water body height (over inland waters) is required. Presently there is an expression for the altitude modelling of the WTC developed by Kouba (2008), however it still has some limitations, namely it should not be used to perform WTC height reductions for heights above 1000 m.
The focus of this study is the modeling of the height dependence of the WTC (or of its symmetric variable, the wet path delay, WPD), aiming to derive improved expressions to account for the complex 4D variation of the WPD. For this purpose, the latest reanalysis model from ECMWF, ERA5, which provides hourly atmospheric parameters at 0.3°x0.3° spatial sampling on 137 vertical levels (from the surface up to around 80 km), is used.
Being ERA5 the first global ECMWF reanalysis model provided at 1-hour intervals, the impact of the different spatial and temporal resolutions of ERA5 and ECMWF operational models is first assessed and shown.
The study is performed in two main steps. First, the errors introduced when using the Kouba expression are assessed. This provides a quantification of the magnitude of the errors and their spatial distribution. Using the knowledge acquired in the first step, improved expressions for the vertical variation of the WPD are determined in the second step.
For use in the first step, two types of WTC are derived from ERA5 at a given level of interest: i) from single-layer parameters at orography level and then reduced to the level of interest with the Kouba expression mentioned above (the so-called 2D approach); ii) from numerical integration of temperature and specific humidity model-level parameters, from the top of atmosphere up to surface and then interpolated to the level of interest (the so-called 3D approach).
A global assessment of the Kouba expression is performed at interpolated altitudes from the Altimeter Corrected Elevations 2 (ACE2) DEM. WTC computed with ERA5 single-layer parameters and then reduced to the altitude of the ACE2 DEM is compared with WTC interpolated from the vertical profiles derived from 3D parameters at the same altitude. This comparison shows WTC differences at DEM altitude with RMS values larger than 3 cm in some land regions of the globe. The origin of these large differences is mainly attributed to errors in the height reduction performed using the Kouba formula.
An independent assessment is performed at GNSS sites by comparing GNSS-derived WTC at station height and those derived from ERA5 using different methodologies (2D and 3D atmospheric parameters) at the same height. Absolute mean and standard deviation values of the WTC differences between GNSS and 3D approach are smaller than the corresponding values of the differences between GNSS and 2D approach. When compared with the 2D approach, the 3D approach can reduce the mean differences between GNSS and ERA5 by a few centimeters.
Secondly, an improved expression for the altitude dependence of the WTC is developed. In this analysis, various levels of complexity are considered, e.g., single expressions with latitude dependence terms and expressions with regional dependence terms. These expressions are crucial for the retrieval of accurate WTC measurements over coastal regions and inland waters, such as rivers and lakes, important to obtain accurate absolute water levels.

Poster show times:

Room Start Date End Date
Foyer, Salao Nobre & tent Thu, Sep 27 2018,18:00 Thu, Sep 27 2018,20:00
Foyer, Salao Nobre & tent Fri, Sep 28 2018,14:00 Fri, Sep 28 2018,15:00
Telmo Vieira
University of Porto, Faculty of Sciences