The 8th WMO Intercomparison of High Quality Radiosonde Systems held in China in July 2010 has produced a large data set on the performance of new operational radiosonde designs, backed up by measurements from Scientific Sounding Instrumentation. This was the result of China supporting this test very actively, and very good collaboration with the various manufacturers.

This allows recommendations as to the radiosonde designs that are potentially suitable for the GRUAN network operations and those which are suitable for routine operations, together with recommendations to improve systems without excessive development expenditure.

At night, most radiosonde systems can provide temperature measurements of suitable quality for both weather and climate work. In the day, many designs need improvement to the sensor exposure to improve the reproducibility of measurements near 10 hPa, and some further testing are needed to reduce the systematic bias between the various radiosondes at these upper levels. With more documentation, several systems have potential for use in GRUAN operations.

The relative humidity sensors tested in Yangjiang had good reproducibility, but several types had large systematic errors at all heights in the troposphere, and the origin of these needs to be identified and rectified as soon as possible. Several systems showed potential for observing relative humidity in the upper troposphere in the tropics, and the new correction schemes seem to have good potential for future observations. It was possible to check the measurements in cloud, using cloud radar (up to 15 km) and ceilometer observations (up to 12 km) to identify some of the clouds. Thus, with more documentation several systems have potential for use in GRUAN operations.

GPS height measurements are very reproducible, and are clearly suitable for all radiosonde operations, given that the equipment is initiated correctly. On the other hand, cheaper radiosondes may be used by exploiting systems such as the Chinese secondary radar for good quality operational results.

Thus, where GPS radiosondes are set up correctly, pressure can be deduced from the geopotential heights and the radiosonde measurements of temperature and relative humidity profile.

As a result of the test, many errors in the various quality radiosonde systems were identified and subsequently rectified ensuring improved accuracy for future radiosondes.

The Radiosonde Comparison was performed as a first collaboration between CIMO and a range of climate scientists associated with GCOS and GRUAN, as part of a WIGOS Pilot Project.

This presentation will provide a summary of the Intercomparison and the main results and recommendations.