COMPARISON OF STRATOSPHERIC TEMPERATURES FROM SEVERAL LIDARS, USING NATIONAL-METEOROLOGICAL-CENTER AND MICROWAVE LIMB SOUNDER DATA AS TRANSFER REFERENCES

Citation
Jd. Wild et al., COMPARISON OF STRATOSPHERIC TEMPERATURES FROM SEVERAL LIDARS, USING NATIONAL-METEOROLOGICAL-CENTER AND MICROWAVE LIMB SOUNDER DATA AS TRANSFER REFERENCES, JOURNAL OF GEOPHYSICAL RESEARCH-ATMOSPHERES, 100(D6), 1995, pp. 11105-11111
Citations number
25
Categorie Soggetti
Metereology & Atmospheric Sciences
Volume
100
Issue
D6
Year of publication
1995
Pages
11105 - 11111
Database
ISI
SICI code
Abstract
Stratospheric temperatures derived from five different lidars are comp ared. Although the lidars are in five separate geographic locations, t he evaluation is accomplished by comparing each of the sets of lidar d ata taken over the course of a year (1991-1992) with temperatures inte rpolated to each location from daily global temperature analyses from the National Meteorological Center (NMC). Average differences between the lidars and NMC temperatures vary for the different lidars by up to 6.7 K. Part of this large average temperature difference is shown to be due to the real temperature variation throughout the day, and the d ifferent times of observation of the NMC data and each of the lidar sy stems. Microwave limb sounder (MLS) data from the upper atmosphere res earch satellite are used to model the diurnal and semidiurnal variatio ns in temperature for each lidar location, for each season. After adju sting for the temperature changes caused by variations in observation time, average temperature differences are reduced among four of the fi ve lidars, compared with the NMC temperatures, but still vary by as mu ch as 3.9 K at stratospheric altitudes between 30 and 45 km. Results o f direct comparisons at two permanent lidar sites with a mobile lidar show that sometimes agreement within 1 to 2 K is achieved, but for oth er cases, larger average differences are seen. Since the precision of lidar temperatures has been estimated to be better than 1 K, further r esearch is needed to reconcile this small expected error with the larg er average differences deduced here using measurements made under oper ational conditions.