The Arm Program's Water Vapor Intensive Observation Periods

Abstract

A series of water vapor intensive observation periods (WVIOPs) were conducted at the Atmospheric Radiation Measurement (ARM) site in Oklahoma between 1996 and 2000. The goals of these WVIOPs are to characterize the accuracy of the operational water vapor observations and to develop techniques to improve the accuracy of these measurements. The initial focus of these experiments was on the lower atmosphere, for which the goal is an absolute accuracy of better than 2% in total column water vapor, corresponding to ~1 W m-2 of infrared radiation at the surface. To complement the operational water vapor instruments during the WVIOPs, additional instrumentation including a scanning Ramanlidar, microwave radiometers, chilled-mirror hygrometers, a differential absorption lidar, and ground-based solar radiometers were deployed at the ARM site. The unique datasets from the 1996, 1997, and 1999 experiments have led to many results, including the discovery and characterization of a large (> 25%) sonde-to-sonde variability in the water vapor profiles from Vaisala RS-80H radiosondes that acts like a height-independent calibration factor error. However, the microwave observations provide a stable reference that can be used to remove a large part of the sonde-to-sondecalibration variability. In situ capacitive water vapor sensors demonstrated agreement within 2% of chilled-mirror hygrometers at the surface and on an instrumented tower. Water vapor profiles retrieved from two Raman lidars,which have both been calibrated to the ARM microwave radiometer, showed agreement to within 5% for all altitudes below 8km during two WVIOPs. The mean agreement of the total precipitable water vapor from different techniques has converged significantly from early analysis that originally showed differences up to 15%. Retrievals of total precipitablewater vapor (PWV) from the ARM microwave radiometer are now found to be only 3% moister than PWV derived from new GPSresults, and about 2% drier than the mean of radiosonde data after a recently defined sonde dry-bias correction is applied. Raman lidar profiles calibrated using tower-mounted chilled-mirror hygrometers confirm the expected sensitivityof microwave radiometer data to water vapor changes, but is drier than the microwave radiometer (MWR) by 0.95 mm for all PWV amounts. However, observations from different collocated microwave radiometers have shown larger differences than expected and attempts to resolve the remaining inconsistencies (in both calibration andforward modeling) are continuing. The paper concludes by outlining the objectives of the recent 2000 WVIOP and the ARM-First International Satellite Cloud Climatology Project (ISCCP) Regional Experiment (FIRE) Water Vapor Experiment (AFWEX), the latter of which switched the focus to characterizing upper-tropospheric humidity measurements.