Deriving meteorological observations from intercepted Mode-S EHS messages. Edmund Keith Stone and Malcolm Kitchen July 28, 2016 Abstract The Met Office has deployed a network of five receivers in the UK which are intercepting routine ADS-B and Mode-S EHS (Enhanced Surveillance) messages broadcast by civil aircraft in response to interrogations by air traffic control (ATC) radars. The information in the messages is used to derive high quality wind and lower quality temperature observations at 4s intervals from most aircraft flying in UK airspace. The total cost of the hardware was only 10k yet the network is currently delivering up to 6.2 million observations per day (including 1000 profiles) or 20 times the volume of upper-air wind observations obtained from all other sources combined. Locating the receivers on weather radar towers ensures a good horizon (hence coverage), and enables sharing of the highbandwidth weather radar communications. Ongoing costs are equivalent to 0.01 per daily wind observation, compared to 25 for AMDAR observations over the UK. Experiments to assimilate these data into UK numerical forecast models are under way and displays of the data for use by weather forecasters have been developed. Global availability of Mode-S data is increasing as ATC radars are progressively modernised and more are requesting the data from aircraft. This trend, along with the potential demonstrated so far with the UK receiver network, have prompted thoughts as to how data coverage could be extended to other regions in future. There seems to be three main possibilities - either deployment of local ground-based receivers, similar to the UK model; establishing a constellation of low-earth-orbit satellites (e.g. cubesats) that include an ADS-B and Mode-S receiver as part of the payload; or obtaining bulk data from ATC authorities around the world. The first two of these options are outlined here. 1 Introduction Observations from commercial aircraft are an important source of atmospheric measurements. With the continued move towards higher resolution modelling it is becoming necessary to also increase the resolution of observations. This is unlikely to be possible using conventional methods of gathering aircraft data 1
to substantially increase the resolution, fortunately promising new methods of sampling the atmosphere using aircraft have begun to be developed [1, 3, 6 8]. These new methods have shown the potential to substantially (more than an order of magnitude) increase the amount of observations that are available. This paper is concerned with the method of deriving wind and temperature data using Mode-Selective Enhanced Surveillance (Mode-S EHS) data. Mode-S is an air traffic control (ATC) standard where aircraft both broadcast data routinely and respond to interrogation from ATC secondary surveillance radar (SSR) with additional information. The Mode-S EHS regime defines a set of interrogation requests that are conducted by the ground SSR in a given airspace. Over the UK most aircraft flying in any airspace will be interrogated reporting their air movement vector (magnetic heading [θ mag ] and true air speed [ v T AS ]) and ground movement vector (as north/south and east/west components). From these two vectors it is possible to derive the wind that is acting on the aircraft [7]. The aircraft also report their Mach number (their speed relative to the local speed of sound) which with the true airspeed can provide a temperature at the location of the aircraft. Significant work has been conducted on accessing the quality of the data this includes cross comparisons with other observations [1, 2, 6], comparisons with numerical weather prediction models, [1, 7] any by synthesising Mode-S data from data on a research aircraft and comparing to the research grade measurements [4]. 2 Local Reception of Mode-S EHS data In the UK there is a network of 5 receivers collecting Mode-S EHS and ADS-B data. These receivers provide between 4 and 6.2 million observations per day over the UK and the surrounding area. As part of the required data is only transmitted on interrogation by ATC SSR it is difficult to access the potential for data in new regions. In Europe most ATC organisations have committed to interrogating aircraft within at least the busiest airspaces before 2020. Beyond this there is very limited information available on where interrogation is happening. It is therefore necessary to conduct a global survey of data availability. The UK network intercepts, processes and populates the operational observations database within 20 minutes of the messages being received by one of the antennas for over 90% of the derived observations. A receiver system consists of a computer, antenna and receiver with the current annual running cost per daily wind observation is 0.01, for UK AMDAR this is 25. The UK data is currently being used for routine monitoring of the data and data assimilation trials, as well as proof of concept tools for forecasters. Figure 1 shows a time series of average wind profiles. Each column represents a ten minute average where the data is split into 100 m high altitude bins and the two wind components are averaged, the newest data is the left hand most column. Figure 1 (a) is the full altitude profile with (b) showing only the lowest portion of the atmosphere. The gaps above 6000 m are due to aircraft tending to manoeuvre at those altitudes on the ascent and descent at Gatwick. 2
Figure 1: Ten minute wind profiles around Gatwick Airport, UK. Each column represents one ten minute, 100 m altitude bins averaged wind. (a) is the full depth of the observations whilst (b) is limited to only the bottom part of the atmosphere. 3 Options for Extending Coverage of Mode-S data 3.1 International Networks of Ground-based Receivers In the UK, the Met Office has chosen to develop a bespoke network of Mode- S receivers specifically for the purpose of meteorological observation. Similar receivers could be deployed in other countries where Mode-s data are available, and either the raw Mode-s data, or the derived observations, could be pooled as for other types of observation. For example, a receiver has recently been installed on the Channel Islands in collaboration with Jersey Met (Jersey Meteorological Department) that streams Mode-s data for processing alongside data from the UK network. The feasibility of creating extensive networks on this model is demonstrated by the networks of mainly amateur ADS-B receivers already exist, see e.g. https://www.flightradar24.com, http://www.adsbexchange.com/. Flightradar 24, with access to data from 10,000 receivers worldwide, provides near total coverage of the global land area. In principle, such networks, could be adapted or replicated in order to relay Mode-S data to national or regional centres for processing, quality control and derivation of meteorological observations. 3
Figure 2: GomX-3, the 3-unit cubesat developed to demonstrate the feasibility of satellite reception and streaming ADS-B data from space 3.2 Reception using a constellation of low-earth-orbit Satellites The European Space Agency sponsored a cubesat mission to assess the feasibility of tracking global aviation traffic using low-cost cubesats [5], see http://www. esa.int/our_activities/space_engineering_technology/tiny_cubesat_tracks_ worldwide_air_traffic. GomX-3 is a 3-unit cubesat developed for ESA by GomSpace (see http://gomspace.com/index.php?p=projects and figure 2). It was launched from the International Space Station in October 2015. The satellite orbits the earth at a height of 400 km with a period of 90 minutes. In July 2016, a follow-on project was initiated by ESA and the Met Office to modify the GomX-3 satellite software to decode Mode-S messages alongside the ADS-B broadcasts. The aims are as follows:- ˆ To demonstrate the feasibility of deriving meteorological data (mainly winds) from Mode-S data received by a cubesat. ˆ To obtain a snapshot of the wind data coverage that could be obtained by a future constellation of satellites with capabilities similar to GomX-3. ˆ To acquire a test dataset that can be used to assist in the development of generalised quality control methods for deriving accurate winds from Mode-S data and applicable anywhere in the world. ˆ To compare satellite and ground-based reception of Mode-s data in terms of coverage and timeliness etc. The satellite is due to re-enter the earth s atmosphere in late 2016 or early 2017, so data collection will be for a limited test period only. If the project is successful, then it suggests that a future constellation of similar cubesats could be used to obtain global Mode-S data in real-time, and when processed, providing upper-air wind data over significant fraction of the global land area. 4 Summary Mode-S is a potentially huge source of meteorological data (mainly winds) that is largely untapped at present. A network of just five receivers in the UK (costing 4
just 2k each) is yielding 4 M high quality wind observations per day that are suitable for NWP assimilation and local use by forecasters. Practical options for extending coverage have been presented that make use of existing technology. If coverage could be extended over most of the worlds land area, then it is projected that 100 m winds per day could be obtained by this means. This figure compares with the 0.6 M winds per day currently being produced by the AMDAR system, References [1] Siebren de Haan. High-resolution wind and temperature observations from aircraft tracked by Mode-S air traffic control radar. Journal of Geophysical Research: Atmospheres (1984 2012), 116(D10), 2011. [2] Siebren de Haan. Estimates of Mode-S EHS aircraft derived wind observation errors using triple collocation. Atmospheric Measurement Techniques Discussions, 8:12633 12661, 2015. [3] Siebren de Haan and Ad Stoffelen. Assimilation of high-resolution Mode-S wind and temperature observations in a regional NWP model for nowcasting applications. Weather and Forecasting, 27(4):918 937, 2012. [4] Andrew K. Mirza, Susan P. Ballard, Sarah L. Dance, Paul Maisey, Gabriel G. Rooney, and Edmund K. Stone. Comparison of aircraft derived observations with in situ research aircraft measurements. Quarterly Journal of the Royal Meteorological Society, pages n/a n/a, 2016. [5] Igor Alonso Portillo1, David Gerhardt, and Morten Bisgaard. Launch and early operations phase for the gomx-3 mission. 2016. [6] Edmund Keith Stone and Malcolm Kitchen. Introducing an approach for extracting temperature from aircraft gnss and pressure altitude reports in ADS-B messages. Journal of Atmospheric and Oceanic Technology, 32(4):736 743, 2015. [7] Edmund Keith Stone and Gary Pearce. A network of mode-s receivers for routine acquisition of aircraft-derived meteorological data. Journal of Atmospheric and Oceanic Technology, 33(4):757 768, 2016. [8] B. Strajnar. Validation of Mode-S meteorological routine air report aircraft observations. Journal of Geophysical Research: Atmospheres, 117(D23), 2012. 5