Non-Nominal Troposphere Reassessment for Meeting CAT II/III with MC/MF GBAS

In Civil Aviation, to meet the long term goal of greater capacity, services must be expanded to provide more reliable, robust approach and landing operations in all weather conditions. This could be achieved globally by using modernized navigation systems. This paper relates to the development of the Multi-Constellation (MC) and Multi-Frequency (MF) Ground Based Augmentation System (GBAS) within the SESAR Framework Work Package 15.3.7. It deals also with the performance improvements obtainable for CAT II/III precision approaches, the most stringent operation currently defined. Several challenges and key issues must be solved including those related to atmospheric modelling. Previous work principally undertaken at Ohio University [1] [2] [3] highlighted the need to consider the troposphere as a possible source of failure. GBAS activities in Europe have followed the approach of validating the protection levels, which includes treating the combined threat relating to ionospheric and tropospheric gradients. Therefore the tropospheric failure should be bounded by validating that the combination of atmospheric errors does not exceed the assumed models. However, there are a number of arguments for revisiting this topic and specifically addressing the tropospheric threat. Firstly, recent observations [4], reported at last ICAO NSP (International Civil Aviation Organisation Navigation System Panel) meeting, showed unexpected atmospheric behavior. The source could be related to a non-modelled behavior of the troposphere. Even if the range errors induced by this phenomenon are not significant compared to those due to ionospheric gradients, the combination of these "tropospheric" gradients with ionospheric gradients could impact integrity and continuity. Secondly, in the advent of dual-frequency GBAS, the ionosphere may feasibly be removed through the Ionosphere-Free (I-Free) smoothing technique. In this case, the main contributor to the atmospheric error will come from the tropospheric delay. Under such a scenario, the troposphere threat model must be defined and a means for bounding the potential errors derived. This paper presents an initial analysis with the aim of evaluating the impact of non-nominal troposphere on VPL for different scenarios. The goal of this comparison is to ascertain the extent to which the proposed tropospheric bounding methodology increases the VPLs used at the aircraft. Finally, this paper has initiated the process of assessing the impact of modelling the non-nominal troposphere on GBAS VPLs. Indeed a new methodology is proposed and seems to improve performance in terms of availability while respecting some constraints on a low data requirements for the VDB transmission.