A43E-0340
A Site Atmospheric State Best Estimate of Temperature for Lauder, New Zealand

Thursday, 17 December 2015
Poster Hall (Moscone South)
Jordis Sonnhild Tradowsky1, Gregory Elton Bodeker1 and Richard Querel2, (1)Bodeker Scientific, Alexandra, New Zealand, (2)NIWA National Institute of Water and Atmospheric Research, Lauder, New Zealand
Abstract:
To detect and attribute changes in local and global climate, long-term observations of the atmosphere are essential. The Global Climate Observing System (GCOS) has defined 50 Essential Climate Variables (ECVs), which are fundamental for understanding the climate system. To fill a long recognized need of the climate monitoring community for reference quality measurements of upper air ECVs, the GCOS Reference Upper Air Network (GRUAN) was established and is now operational.
This study presents a temporally highly-resolved Site Atmospheric State Best Estimate of temperature (T-SASBE) for the GRUAN site at Lauder, New Zealand.
In a SASBE all available measurements of a selected ECV at one specific site (and possibly including estimates obtained from measurements made elsewhere) are merged to produce a best estimate of the value of that ECV and the uncertainty on each datum.
The first version of T-SASBE, extending from 1996 to 2006, combines measurements from: (i) weekly radiosondes launched at Lauder, (ii) 10-minute data obtained from the surface weather station at Lauder and (iii) twice-daily radiosondes launched at Invercargill, about 200km south-west of Lauder. Later versions of T-SASBE will extend the time period and include further measurements made at Lauder.
The temperature profiles measured at Invercargill are included in T-SASBE to provide an estimate of the vertical structure in temperature at times between the once-weekly radiosonde flights at Lauder. Corrections to the Invercargill measurements were derived by applying a regression model to temperature profiles measured at Lauder and Invercargill within 12 hours of each other. Different predictor variables, including wind speed and surface pressure, were tested and the quality of the regression model was evaluated. The regression model with the best combination of the predictor variables was used to correct the Invercargill temperature profiles and use them as a valid proxy for 12-hourly temperature profiles at Lauder.
While the version of T-SASBE presented here is the first milestone in the process of building the T-SASBE above the GRUAN site at Lauder, further improvements to be implemented in the future are expected to provide a SASBE that includes robust estimates of the uncertainty suitable for e.g. validating space-based measurements over Lauder.