Report on the Status of Interoperable Upper-Air Systems

WORLD METEOROLOGICAL ORGANIZATION

COMMISSION FOR INSTRUMENTS AND METHODS OF OBSERVATION CIMO/MG-3/ Doc. 3.4.3 (7.VI.2006) CIMO MANAGEMENT GROUP Third Session ______ITEM: 3.4 GENEVA (SWITZERLAND), 3-7 JULY 2006 Original: ENGLISH ONLY

REPORT ON THE STATUS OF INTEROPERABLE UPPER-AIR SYSTEMS

(Submitted by Hydro-Meteorological Equipment Industry Association, Mr R. Dombrowsky, OPAG-UPPER-AIR Co-Chair, ET-UGRN Chair, Mr J. Nash, OPAG-UPPER-AIR Co-Chair, ET-UASI Chair) ______

Summary and purpose of document

This document summarizes the outcome of the study by ET-UGRN and HMEI on the Interoperable Upper-Air Systems.

Action proposed

The MG is invited to consider the information provided in this document, review and endorse appropriate proposals and recommendations for the deliberation by CIMO-XIV. CIMO/MG-3/ Doc. 3.4.3, page 2

EVALUATING INTEROPERABILITY IN UPPER AIR SOUNDING SYSTEMS

Frederick A. Clowney International Met Systems, Grand Rapids, Michigan (on behalf of HMEI)

1. Introduction 3. History The WMO Executive Council has requested that Upper-air sounding systems consist of two the Secretary General investigate the feasibility primary components: a radiosonde that of interoperability in upper-air sounding systems. measures and transmits meteorological data and This topic was first addressed in a paper a ground station that receives the telemetry and delivered to the WMO Technical Conference on processes it into met data products. Meteorological and Environmental Instruments The earliest ground stations were optical and Methods of Observation (TECO 2005, theodolites that tracked pilot balloons to Bucharest).1 determine winds. Interoperability was an A second paper was delivered to the obvious attribute since there were no electronics Commission for Instruments and Methods of involved and any balloon could be used. Observation Expert Team on Upper-Air System When automatic sounding systems were Intercomparisons, Second Session (CIMO ET on introduced along with electronic radiosondes, UASI 2005, Geneva).2 The CIMO paper interoperability was generally maintained. An addressed technical issues presented by example of this is the ART system installed by interoperability in both 1680 and 403 MHz the U.S. National Weather Service (NWS) in the systems. 1950s. The purpose of this paper is to summarize the The introduction of systems using early findings of the earlier papers and present a set navigational aids (Loran-C, OMEGA) began to of options for CIMO-XIV in December 2006. reduce interoperability. The introduction of 403 2. Definitions MHz GPS systems in the 1990s effectively eliminated it.3 “Interoperability” is defined as the ability of an upper-air ground station to employ radiosondes Interoperability disappeared in GPS-type made by multiple manufacturers, including, but systems for two reasons. The first is that most not limited to, the manufacturer of the ground of the early systems used proprietary GPS station. coding schemes that required the ground station to be highly integrated with the radiosonde. The From the operators’ perspective, the goal of second is that the relatively low cost of GPS interoperability is to reduce costs while ground stations appeared to make maintaining quality in upper-air soundings. This interoperability less important. If there were would be accomplished by introducing greater problems with the radiosonde supply, the ground competition into radiosonde re-supply tenders station could be easily replaced. Interoperability also gives operators the ability to The weakness of the second argument became select a new sonde if a more technologically clear towards the end of the decade when some advanced device becomes available. This stations began to go dark because they could would prevent operators from becoming locked not afford the high cost of GPS operations. into an obsolete sonde due to a ground station that is incompatible with newer disposables. 4. Economics The trend in upper-air systems has been to move costs from the ground station to the Corresponding author’s address: Frederick A. radiosonde. This is most easily seen by th Clowney, InterMet Systems, 4460 40 St SE, comparing radiotheodolites, where a relatively Grand Rapids, MI 49512; e-mail: high cost ground station is used with a relatively [email protected] low cost radiosonde, to GPS systems (where the opposite is true.)4 In a typical GPS installation, CIMO/MG-3/ Doc. 3.4.3, page 3 the cost of disposables will exceed the cost of the ground station in the first year or operations.5 Radiosonde Cost Trends When GPS sondes were first introduced, the cost of fully coded GPS receivers was quite high. This led some manufacturers to use non- coded devices that were less expensive but suffered from poor performance. Current generation 403 MHz sondes use fully-coded GPS receivers, taking advantage of falling costs Figure 1 : RDF Upper Air Sounding System for these devices. Although this has improved performance it has not resulted in significant declines in radiosonde prices. Based on materials costs alone, there is no longer a clear reason why GPS sondes should be more expensive than current generation RDF devices. In practice, GPS sondes remain 35% to 100% more expensive than their RDF counterparts. The most obvious reason for this discrepancy is the lack of competition for GPS radiosonde re- Figure 2 : GPS Upper Air Sounding System orders.6 This is supported by the observation that the most intensive price competition comes An important difference between the two formats when a ground station is being replaced and is that the 1680 MHz receiver is located in the there is an incentive for manufacturers to offer antenna for RDF systems. 403 MHz GPS the lowest possible sonde prices. After the initial systems require two receivers (UHF and ground station order, competition effectively differential GPS) both of which are located in the stops, providing no incentive to maintain Meteorological Processor (MP). This makes the competitive prices. MP a much more complex and costly device than the corresponding signal processor used in If this is in fact the case, interoperability could RDF systems. benefit operators by introducing competition into radiosonde re-supply tenders. Plug-and-Play Systems 5. Technical Considerations If upper-air systems were analogous to PCs, interoperability would be the equivalent of “plug- Upper-air sounding systems consist of five main and-play” devices. If upper-air systems were elements: interoperable, a radiosonde would be a device - Radiosonde / Transmitter that would be easily recognized by the other - Antenna(s) / Receiver(s) parts of the system. If the sonde was new and the system didn’t recognize it, the software - Signal Processing System (decoder) would request the necessary “drivers” to allow - System Computer for use of the new device. - Meteorological Operating System (software) As upper-air systems have developed, In addition to the five main elements, there may manufacturers have moved away from the plug- be peripheral equipment specific to certain and-play concept and created closed systems manufacturers such as radiosonde ground check that only accept components made by the same devices. manufacturer. In the PC world, this would be the same as Dell saying that its PCs can only use Figure 1 and Figure 2 show simplified system Dell printers, Dell disc drives, Dell monitors, etc. diagrams for RDF and GPS type systems. There are several reasons why upper-air systems have become closed rather than open, plug-and-play type systems:

3 CIMO/MG-3/ Doc. 3.4.3, page 4

- Manufacturers use proprietary methods to Processor used in GPS systems (Figure 2) decode, correct and process the PTU data is significantly more expensive since it collected by their radiosondes. These includes the system’s receivers as well as methods cannot be disclosed without putting the sonde decoder. trade secrets at risk.  The UHF antennas and Low Noise - Plug-and-play compatibility is expensive to Amplifiers (LNAs) included with 403 MHz develop and there is no incentive for GPS systems are not standardized and manufacturers to provide it. need to be carefully integrated with the corresponding receivers in the MP. - Manufacturers have an incentive to control all parts of the system in order to maintain  Algorithms required for the Meteorological quality and provide seamless integration. If Operating System are more extensive than one manufacturer does not control the just calibration and solar correction. Since complete system, it becomes difficult to most sonde manufacturers use proprietary determine who is responsible if there is a GPS wind finding schemes, this code would system malfunction. also have to be integrated. - Users have generally not required it. The worst case situation for a 403 MHz GPS system would require over 90% of the ground Interoperability in RDF systems. station to be replaced for it to be compatible with 1680 MHz RDF systems have demonstrated the a new sonde. This obviously does not qualify as technical feasibility of interoperability. For an interoperability since you are essentially getting RDF system to use a new radiosonde, two a new ground station. conditions have to be met by the sonde For 403 MHz GPS systems to be operationally manufacturer: interoperable MPs would have to be 6. A sonde-specific Signal Processing System standardized to allow for the swapping of has to be supplied that is compatible with the decoder cards only, while maintaining the antenna and system computer receivers, LNAs, etc. This would allow the ground station to use the same receivers and 7. Certain algorithms have to be provided to antennas while changing only the sonde-specific the antenna supplier so the Meteorological element. Achieving this would require an Operating System can apply sonde-specific agreement among the manufacturers for calibration and data corrections. standardized board sizes and pin-outs. This After a new sonde has been integrated into the could require re-designing existing MPs, which operating system, changeover from one sonde would involve significant costs. to another should be possible in a matter of 8. Conclusions minutes. Interoperability has been demonstrated in 1680 Interoperability in GPS systems MHz RDF systems in synoptic operations. Although it is theoretically possible, operational Interoperability in 403 MHz GPS systems will interoperability has not been demonstrated in only become viable if the user community 403 MHz GPS systems. There are three main demands it from the upper-air system reasons for this: manufacturers. This would require increased  The Signal Processing System that is standardization in ground stations and a swapped out for compatibility in RDF movement away from closed systems into systems (Figure 1) is a relatively simple and something more akin to “plug-and-play”. low cost device. The Meteorological

4 CIMO/MG-3/ Doc. 3.4.3, page 5

End Notes

5 1 “Universal Upper Air Sounding System”, Rodney D Wierenga Ph.D., Frederick A. Clowney, International Met Systems, 2005 2 “Interoperability in Upper Air Sounding Systems” Frederick A. Clowney, International Met Systems, 2005 3 The one exception to this statement is the Telemetry Receiving System (TRS) developed by the NWS in 2001, which required interoperability in a 1680 MHz GPS system. The NWS is the only major weather service that has adopted this approach. 4 Wierenga et al, op. cit. pp 5-6 5 Assumes $50,000 in capital costs for a station flying one sonde per day at a cost of $150 per sonde, or $54,750 per year. 6 Another explanation for the price discrepancy between RDF and GPS sondes is the relatively old age of RDF designs available in the market. This has allowed manufacturers to become more efficient in building them and has also eliminated most, if not all amortization of design costs. An RDF sonde meeting the highest international performance levels is likely to be more expensive than the legacy models currently being sold.