The Impact of Severe Ionospheric Conditions on the Accuracy of RTK Position Estimation: Performance Analysis of Various Ionospheric Modeling Techniques

Dorota A. Grejner-Brzezinska, Pawel Wielgosz and Israel Kashani, Dru A. Smith, Paul S. J. Spencer, Douglas S. Robertson, Gerald L. Mader, Attila Komjathy

Abstract:	"Real-time kinematic (RTK) GPS positioning is subject to well-known limitations resulting from the spatially correlated effects, such as differential atmospheric or orbital errors that increase with the growing distance between the base and the rover stations. This may ultimately prevent the integer ambiguity resolution (AR), and may result in lower positioning accuracy. In order to reduce the impact of the atmospheric errors, decorrelating with the growing spatial base-rover separation, continuously operating global and regional reference GPS networks have been established. The dual frequency realtime data from these GPS arrays can be used to estimate the ionospheric and/or tropospheric errors, since the reference station coordinates within the arrays are known. The estimated atmospheric corrections are broadcast to the users, and used to correct the user observational data, leading ultimately to more reliable and faster AR and increased accuracy of the user position. The accuracy, as well as the time and space resolutions of the atmospheric models, are crucial to reliable AR and rover positioning. Namely, too low time/space resolution or biases in the ionospheric product will lead to errors in AR and rover positioning. In this paper, five methods of the ionospheric correction modeling and estimation, with varying spatial and temporal resolutions, based on GPS permanently tracking reference networks, are presented. The methods are compared based on their spatial extent, sampling rate (i.e., correction update rate) and method of ionosphere modeling. The ionospheric reference ÒtruthÓ is generated in the form of double-differenced (DD) ionospheric corrections, estimated directly from the dual frequency carrier phase data using fixed DD ambiguities. The primary algorithmic and software tool is the Multi Purpose GPS Processing Software (MPGPSª), developed at The Ohio State University. The primary objective is to assess the quality of each model and its applicability to AR and rover positioning. In this paper, the focus is on severe ionospheric conditions, while the earlier publications (Grejner-Brzezinska et al., 2004 and 2005) provided detail analyses of the impact of various ionospheric models on AR and rover positioning under benign ionosphere. This paper provides an assessment of the effects of severe ionospheric storms on GPS observations, and analyzes the quality and reliability of the externally provided ionospheric models supporting AR and kinematic positioning. The ionospheric models that will be tested are: (1) absolute (biased) carrier phase-based model, decomposed from double-differenced (DD) ionospheric delays (OSU MGPS-NR model), (2) tomographic model using pseudorange-leveled L1-L2 phase data (NGS MAGIC model), (3) NGS MAGIC model derived in near real-time (MAGIC-NRT), (4) code and carrier phasebased JPL global ionospheric maps (GIM), and (5) International GNSS Service (IGS) GIM. Method 1 assumes that the ionosphere is an infinitesimal single layer, while methods 2 and 3 considers the ionosphere as a 3D medium; model 4 is based on a set of local horizontal basis function representation of the global ionosphere, and model 5 is a combination of several solutions derived independently by the IGS Analysis Centers. Model 1 is a local ionosphere representation available at the data sampling rate, methods 2 and 3 are regional models (at 15-minute intervals), while models 4 (time resolution up to 5 minutes, 1-hour resolution was tested) and 5 are global (2-hour time resolution). A 24-hour data set collected by the Ohio Continuously Operating Reference Station (CORS) network on November 20, 2003 (Kp index reaching up to 8+) is presented and analyzed, with a special emphasis on the varying ionospheric conditions during the benign and stormy intervals through the 24-hour period. Reference ionosphere is generated first, and the ionospheric models are assessed with respect to the reference truth. One CORS station is selected as rover, and the data reduction is performed in the post-processing mode. The impact of these models on the speed and reliability of the AR process and the positioning accuracy at the user location is discussed. In particular, the time required to fix the ambiguities (i.e., time-to-fix) applying various ionospheric models, and the quality of the resulting rover position coordinates as a function of ionospheric conditions are presented."
Published in:	Proceedings of the 61st Annual Meeting of The Institute of Navigation (2005) June 27 - 29, 2005 Royal Sonesta Hotel Cambridge, MA
Pages:	887 - 901
Cite this article:	Grejner-Brzezinska, Dorota A., Wielgosz, Pawel, Kashani, Israel, Smith, Dru A., Spencer, Paul S. J., Robertson, Douglas S., Mader, Gerald L., Komjathy, Attila, "The Impact of Severe Ionospheric Conditions on the Accuracy of RTK Position Estimation: Performance Analysis of Various Ionospheric Modeling Techniques," Proceedings of the 61st Annual Meeting of The Institute of Navigation (2005), Cambridge, MA, June 2005, pp. 887-901.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In