Radiosonde upperair measurements

Upper-air climatology is less well known than surface conditions, but upper-air observations provide direct measurements of atmospheric state variables above the Earth's surface. These measurements are an important complementary data source for understanding atmospheric circulation and its transport of energy and mass. The upper-air data provide a vertical dimension that reveals the atmospheric response to time and space variations required to maintain energy, mass, and momentum balances and the circulation patterns shaped by these adjustments. The circulation patterns are major factors in atmospheric moisture transport that is fundamental to climate of the first kind and the atmospheric branch of the hydrologic cycle.

Measuring vertical profiles of atmospheric state variables provides understanding of how the atmosphere responds to and contributes to conditions observed at the Earth's surface. The collection of data in the free atmosphere relies heavily on radiosondes. A radiosondes is a small, lightweight package of battery-powered, expendable meteorological instruments attached to a helium-filled free-flying balloon. The balloon-borne instruments simultaneously measure and transmit meteorological data while ascending through the atmosphere at intervals that vary from 1 to 6 s depending on the type and manufacturer of the radiosondes (Dabberdt et al., 2003). When wind information is processed by tracking the balloon's movement the instrument package is known as a rawinsonde.

Radiosondes provide vertical profiles of atmospheric temperature, pressure, humidity, and wind from the Earth's surface up to a height of 20 to 30 km where atmospheric pressure is 10 hPa (Fig. 3.5). The data are transmitted in a predetermined sequence to a ground receiving station where the data are processed at a fixed time interval. Radiosonde temperature data are supplemented by aircraft observations, particularly over the oceans where radiosonde launches are limited to remote islands or ships (Dabberdt et al., 2003). Combining radiosonde and aircraft data requires consideration that the aircraft may disturb the temperature, pressure, and moisture fields it is measuring to a greater extent than a balloon.

The global radiosonde network includes about 900 upper-air stations and approximately two-thirds of these stations make observations twice daily. Global radiosonde launchings at 0000 and 1200 Coordinated Universal Time (UTC) began in 1957. WMO member countries form part of the Global Observing System of the World Weather Watch program and share their sounding data with other members. The synoptic radiosonde/rawinsonde observing programs are designed to meet real-time operational needs for weather forecasting and analysis requiring simultaneously acquired observations at a large number of locations and with a high vertical resolution. The radiosonde network is predominantly land-based and favors the middle latitudes of the Northern Hemisphere. Unfortunately, there has been a deterioration of the radiosonde network due to the loss in September 1997 of the Omega radionavigational system used to track the sondes and the closure of radiosonde stations in some countries to reduce operating costs. The eight-station Omega system was a global network, but it became too expensive to operate with the success of Global Positioning System (GPS) technology.

Global Radiosonde Graph

Fig. 3.5. Radiosonde-based atmospheric sounding for 0000 UTC 20 November 2006 at Oakland, California (38° N, 122° W). The bold line plotted to the right is atmospheric temperature, and the bold line plotted to the left is dew point temperature. Pressure coordinates (hPa) are used for the vertical axis. Wind direction and speed are shown along the right margin with the barb pointing into the direction from which the wind blows and the length of the flags indicating wind speed. (Data courtesy of the NOAA Forecast System Laboratory from their website at

Fig. 3.5. Radiosonde-based atmospheric sounding for 0000 UTC 20 November 2006 at Oakland, California (38° N, 122° W). The bold line plotted to the right is atmospheric temperature, and the bold line plotted to the left is dew point temperature. Pressure coordinates (hPa) are used for the vertical axis. Wind direction and speed are shown along the right margin with the barb pointing into the direction from which the wind blows and the length of the flags indicating wind speed. (Data courtesy of the NOAA Forecast System Laboratory from their website at

3.7.1 Radiosonde instruments

There is no worldwide standard radiosonde, and radiosonde observations are made with a greater variety of sensors and devices than surface observations (Elliott et al., 2002). The most widely used radiosondes are produced by China, Finland, Japan, Russia, and the United States (Luers and Eskridge, 1998). In general, the instruments in these radiosondes have different error and response characteristics and undergo modifications and improvements during the life-time of a particular model.

Temperature sensors in contemporary instrument packages are thermistors, capacitive sensors, resistance wires, bimetallic elements, or thermocouples (WMO, 1996). Radiosondes widely used in the United Kingdom and the United States measure temperature with a temperature-sensitive capacitor consisting of two electrodes separated by a ceramic dielectric (Hudson et al., 2004). Atmospheric temperature measurement by every radiosonde instrument is affected by heating from sources other than the air itself that make the sensor temperature different from the ambient air temperature. A temperature correction is required to determine the actual air temperature. This correction is small at altitudes below 15 km but can be substantial at altitudes between 20 and 30 km. A major difference among all radiosondes is the nature of the temperature correction applied to the temperature sensor value (Eskridge et al., 2003).

The most widely used radiosondes determine atmospheric pressure changes using the change in measured capacitance that occurs between two capacitance plates (Dabberdt et al., 2003). Other radiosondes use a classical aneroid-type sensor that is two electrodes separated by a distance that varies depending on the volume of a partially evacuated and expandable cell. A piezoresistance element is sometimes used for pressure measurements, and a water hypsometer has been used successfully by the Swiss Meteorological Institute (Richner et al., 1996).

Most radiosonde instrument packages have an electric resistive hygrometer to determine humidity. This device uses a thin piece of lithium, a carbon hygristor, or other semiconductor material and measures the resistance which is affected by humidity. An alternative design is two electrodes separated by a thin polymer film. The capacitance varies depending on the amount of water absorbed by the polymer film and on the film temperature. Relative humidity is calculated from the capacitance and temperature data (Hudson et al., 2004).

A radio transmitter sends the sensor data to a ground-based receiving station. The balloon ascends at a rate of 5 to 8 m s"1 until it eventually bursts due to the low pressure and low temperature. This rate of ascent allows the measurements to 30 km to be completed in about 90 minutes. The instrument package may be equipped with a parachute for its descent back to the surface, but only 25% of routinely released radiosondes are ever recovered (DeFelice, 1998).

Wind is not measured directly by the radiosonde, but wind directions and velocities are determined from the balloon drift. Determining the balloon drift requires the ability to track the balloon after it is launched or to locate the balloon during its flight. Balloon tracking uses an optical system, a radio signal, or radar. The optical system uses a theodolite to determine the balloon's azimuth and elevation. The radio signal uses a radio theodolite to determine the same information from the transmitter on the balloon. Both of these methods require knowledge of the balloon's height which is determined from the pressure measurement, and both require clear conditions between the balloon and the ground station. Equipping the balloon-borne package with a radar reflection shield permits the balloon's position to be tracked by radar. Tracking changes in the balloon's position by three or more ground-based stations provides a basis for calculating wind directions and velocities. All three of the tracking methods have reduced accuracy as the distance to the balloon increases and its elevation angle decreases (Dabberdt et al., 2003).

Locating the balloon during its flight is achieved using one of the navigation systems developed for other purposes. The LORAN-C is a radionavigation system developed by the United States, Canada, and Russia for maritime purposes. A transponder on the balloon receives the navigation signals and rebroadcasts them back to the ground station with the meteorological measurements. This system requires pressure measurements to determine the height of the balloon. An emerging navigation-based technique uses the satellite-based GPS. This technique is very accurate for measuring the position and velocity of the radiosonde, and the GPS has worldwide coverage. However, the cost of GPS technology is high for an expendable application (Dabberdt et al., 2003).

3.7.2 Radiosonde data

Archived radiosonde data in national and international databases begin after 1945 and provide upper-air observations to support a variety of operational and research applications important to hydroclimatology. Spatial patterns of temperature and humidity and pressure and temperature trends at the surface and in the troposphere and lower stratosphere have been developed from radiosonde data (Dai et al., 2002; Angell, 2003). Constructing temperature and water vapor climatologies and parameterizing water vapor and cloud processes are other areas employing radiosonde data (Ross and Elliott, 2001; Miloshevich et al., 2004). However, utilizing archived radiosonde data for hydroclimatic research requires caution due to the variety of sensors employed globally and changes in instrumentation and observation methods that introduce biases in the data (Elliott et al., 2002; Angell, 2003). Also, radiosonde data are subject to errors due to radiational heating and cooling of the sensor arm, calibration errors, and chemical contamination of the sensors (Hudson, et al., 2004). Removal of errors and biases must precede use of the data for hydroclimatic analysis. Metadata, satellite data, and heat transfer models of the instruments are employed to identify and remove errors and biases (Parker et al., 2000).

Each balloon ascent, or sounding, produces about 20 data levels with readings on temperature, pressure, humidity, and related winds. With two soundings each day reported for as many as 2300 upper air land and marine stations, the magnitude of the raw data is an immense validation and data management problem. The WMO's GCOS establishes the specific record length and homogeneity requirements for upper-air stations. The GCOS Upper Air Network (GUAN) includes the globally distributed stations that meet the GCOS requirements. GUAN strives to ensure the maintenance of key stations and availability of their data (Parker et al., 2000). Quality control and archiving of upper-air data from the GUAN stations is jointly managed by the Hadley Centre of the UK Meteorological Office and NOAA's National Climatic Data Center (NCDC).

Two sets of radiosonde data for GUAN stations are available. Current data from the Global Telecommunications System (GTS) consist of soundings received at NCDC that receive no quality control screening other than review for duplicate records and syntax errors. Historical upper-air data are archived in the Comprehensive Aerological Reference Data Set (CARDS) which is a joint project of NOAA's NCDC and the All-Union Research Institute of HydroMeteorological Information, Russia. CARDS data consist of soundings from multiple data sources that have been processed by the CARDS Complex Quality-Control system (Eskridge et al., 1995). Additional adjustments to the CARDS data to remove inhomogeneities are described by Free et al. (2002). CARDS contains over 27 million quality controlled radiosonde observations for the period 1940 to 2000, and this number increases with each revision of the CARDS dataset. Monthly mean data (MONADS) for each CARD station are available (Elliott et al., 2002). In addition, data sets of monthly and seasonal temperature anomalies and monthly CLIMAT TEMP are provided by the Hadley Centre Radiosonde Temperature (HadRT) products based on monthly average temperature reports from radiosonde operators.

Continue reading here: Review questions

Was this article helpful?

+1 0