Skip to content

*Glidersonde Feasibility Report

A Demonstration of the Feasibility of the Glidersonde for the
Deployment and Recovery of Weather Instrumentation

Davis M. Egle and Michael L. Babb
University of Oklahoma
Norman, Oklahoma

1. Introduction

(download in pdf format, 13K)

1.1 Background and Need

The accuracy of the weather prediction models, and the impact that accurate forecasts can have on national economic activities, depends on the distribution and frequency of sampling of atmospheric data. Presently the ubiquitous and "expendable" radiosonde package is the instrument of choice, at least in most developed countries. They are also used in developing countries, but the cost of the units has prevented widespread use. Radiosonde observations are considered the single most important component of the current atmospheric observation system, providing essential information required for weather prediction models.

The most popular method for employing a radiosonde is to attach it to a helium or hydrogen filled weather balloon, launch it, and monitor the radio transmissions until the balloon bursts or goes out of radio range. The data sent normally includes pressure, temperature, and humidity. When the balloon bursts the instrument package, which has a mass of about 0.3 kg, falls to earth. The difficulty and cost of recovering the instrument normally exceeds the value of the device so they are generally considered expendable.

There are a variety of techniques for wind determination from radiosonde systems. Most radiosonde systems in the US use a radiotheodolite, a parabolic antenna that automatically follows the radiosonde by pointing in the direction of the maximum signal strength. However, because of low maintenance costs, lower initial costs, and simplicity of the equipment, a Global Positioning System (GPS)-based windfinding system is being adopted in many countries. Radiosonde manufacturers have responded to this change by developing a radiosonde based on the GPS. The cost of the GPS sondes however is higher, and is near $160 a unit, compared with $90 per unit for older systems. This in effect almost doubles the cost of providing the atmospheric data needed for the forecast prediction models.

The use of expendable radiosondes worldwide is high. More than 1000 expendable sondes are launched daily with about 150 of these being launched in the US alone. Thus the total daily expenditures worldwide is about $100,000 with $15,000 a day being spent in the US. Should a full conversion to GPS systems occur, the daily costs will increase to approximately $200,000 worldwide and $25,000 in the US. This is a significant cost increase for many weather services to absorb.

1.2 Goal

The goal of this project was to demonstrate a cost-effective method for recovering the radiosondes used in measuring atmospheric data for weather prediction models. The availability of a recoverable sonde will allow an increased distribution of the device and an increased frequency of sampling. A recoverable instrument package will result in greater accuracy of the prediction models or in lower cost for the current accuracy of forecasts or a combination of both.

The ideal solution would be a reusable vehicle that is launched in a manner consistent with current practice (i.e. by balloon) but would return to the launch site or to another recovery site. This would allow the reuse of the vehicle and the instrumentation and a subsequent significant reduction in the cost of the operation. To provide an expanded distribution, the ideal device would need to be designed to the lowest possible acquisition cost. This concept will provide the weather service with a flexible means to balance the trade-off between the accuracy of prediction and the cost constraints for daily predictions and research efforts. Additionally, given the large numbers of devices launched daily, the negative environmental impact of the disposable devices would be eliminated.

1.3 Report Overview

The remainder of this report addresses the design philosophy - the mission, aerodynamic design, and vehicle design of a Glidersonde; the design of a prototype vehicle for demonstrating the feasibility of the concept; the navigation system; and the flight tests conducted with the prototype Glidersonde.

2. Glidersonde Design

(download in pdf format, 113K)

Overview and technical description of Glidersonde design criteria in general.

3. Prototype Design

(download in pdf format, 40K)

Design details for the Glidersonde prototypes used in the test flights.

4. Navigation System

(download in pdf format, 152K)

This chapter describes the computer algorithms that were used in the navigation computer developed for the Glidersonde.  This chapter also describes the computer interfaces with the Global Positioning System (GPS) receiver, control servos, and the human operator.

5. Flight Tests

(download in pdf format, 60K)

Discussion of the Glidersonde flight tests that have taken place as of May 1999.

6. Conclusions

(download in pdf format, 6K)

The goal of this project was to demonstrate the feasibility of the Glidersonde concept.  Toward that goal, a design approach was developed and shown viable through the construction of a simple prototype vehicle and the subsequent flight tests.  The developed navigation system was shown capable of separating the vehicle from the launch balloon and successfully navigating the glider back to the home location.  The prototype vehicle was shown to return successfully in average winds up to 21 knots.

A low-cost telemetry system was developed and shown to function effectively in the real-time monitoring the position, altitude, heading, and speed of the glidersonde.  The telemetry system is considered a necessary element in the glidersonde instrumentation.  The meteorological instrumentation could easily be integrated into the telemetry system to further reduce costs of the system.

The achievement of altitudes up to 20 or 25 thousand feet should be relatively simple with the current system.  Higher altitude will require more development because of the low temperatures encountered at the higher altitudes and the use of a higher speed vehicle because of the high-speed winds encountered at altitudes above 25000 ft.  However the design approach developed in this report indicate that such a vehicle is well within the limits of existing technology.


Appendix A  (231K pdf format)
Screen shots of the test flight paths for the flights described in section 5.

Appendix B  (669 K pdf format)
Plots of altitude vs. time, speed vs. time and speed vs. heading for each of the test flights described in section 5.


(download in pdf format, 8K)

1. RADIOSONDES -- An Upper Air Probe Offsite link warning by Edward J. Hopkins, June 1996

2. Lanzante, J. R., and G. E. Gahrs, 1997: Examination of some biases in satellite and radiosonde measures of upper tropospheric humidity using a framework for the comparison of redundant measurement systems. In Proceedings of the Twenty-First Annual Climate Diagnostics and Prediction Workshop, Springfield, VA: NTIS, 352-355.

3. Lanzante, J. R., 1996: Resistant, robust & non-parametric techniques for the analysis of climate data: Theory and examples, including applications to historical radiosonde station data. International Journal of Climatology, 16(11), 1197-1226.

4. McPherson, R. M., 1999: The future of the North American radiosonde network.  Third Symposium on Integrated Observing Systems, 10-15 January, 1999.  Dallas, Texas. pp 14-17.

5. Lalley, V.E., 1991:  A reference radiosonde.  Seventh Symposium on Meteorological Observations and Instrumentation.  Jan 14-18, 1991.  New Orleans, LA.  pp 217-220.

6. Maselli, Brian P., 1998: A Preliminary Design Study of an Autonomous Glidersonde, Master's Thesis, School of Aerospace and Mechanical Engineering, University of Oklahoma, 36 pgs.

7. McCormick, Barnes W.: Aerodynamics, Aeronautics, and Flight Mechanics, John Wiley & Sons, New York, 1979, 652 pgs.

8. Wood, K.D.: Arospace Vehicle Design ? Volume I Aircraft Design, 3rd Edition, Johnson Publishing Company, Bouylder, Colorado, 1968


(download in pdf format, 5K)

The authors gratefully acknowledge the contributions of several people to this project.  Dr. Dudley Smith and Mr. Brian Maselli contributed to the early work which culminated in Mr. Maselli's Master Thesis which in referenced in this report.  Dr. Frank Gallagher, Assistant Professor in Meteorology at the University of Oklahoma, contributed to the early phases of the telemetry system development.

Ken Howard and Dr. Mike Douglas, of the National Severe Storms Laboratory, Norman, OK, contributed significantly to the success of this project in ideas and suggestions, help in the flight tests, especially the balloon tests, and through their support for the financing of the project.  Finally we appreciate the support of Dr. James Kimpel, Director of NSSL, Norman, for his support of the project.