CoAuthors

Denise Dettmering (Deutsches Geodätisches Forschungsinstitut der Technischen Universität München (DGFI-TUM), Germany); Michael Schmidt (Deutsches Geodätisches Forschungsinstitut der Technischen Universität München (DGFI-TUM), Germany); Andreas Goss (Deutsches Geodätisches Forschungsinstitut der Technischen Universität München (DGFI-TUM), Germany)

Recently, the demands for atmospheric products (e.g., electron and neutral density), useful as input to real-time applications (e.g., telecommunication, navigation and positioning, satellite mission control, solar event monitoring and forecasting), have been rapidly increasing. In this context, international agencies such as the International GNSS Service (IGS) have been spending efforts to provide raw data with low latency from space-geodetic techniques to support real-time and near real-time (NRT) modelling approaches, e.g., the ionospheric Vertical Total Electron Content (VTEC) modeling. The IGS, for instance, provides global terrestrial GNSS data through its data centers in RINEX format with daily and hourly latencies as well as real time products via Ntrip data streams.

GNSS observations cover the continental regions and due to their spatiotemporal resolution have very high importance for studying VTEC variations. However, large data gaps exist due to the inhomogeneous distribution of the GNSS observation sites, especially over the oceans. Therefore, other techniques such as DORIS, e.g. on board of the satellites Jason-3 and Sentinel-3, can mitigate the data gap problem and improve the accuracy and
reliability of ionospheric maps. Furthermore, DORIS can contribute to a data densification in continental regions.
However, data derived from the DORIS system is currently provided with a latency of several days – too late for rewarding an inclusion in (near) real-time ionospheric models. In the frame of the IDS working group on “NRT DORIS data”, our study is devoted to investigate the contribution of NRT DORIS data to ionosphere modelling.

At present, DGFI-TUM’s VTEC model is mainly supported by GNSS hourly data. It is based on a series expansion in two-dimensional B-spline functions constructed as tensor products of two one-dimensional basis functions, namely trigonometric and polynomial B-splines. All observations are exploited in a Kalman filter to estimate the unknown ionospheric target parameters, namely the B-spline coefficients, differential code biases, etc.

In this contribution, we simulate the case that low-latency DORIS data exist and study their impact on the quality of NRT VTEC products. To be more specific, we generate DORIS data of different latencies from 1 hour to 6 hours and include them in our VTEC model by using a multi-filter approach. The different model performances are assessed and compared to the results of the current GNSS-only approach.