An intercomparison of precipitable water vapor data from satellite and groundbased remote sensing devices

The limited application of the Television and Infrared Observational Satellite (TIROS) Operational Vertical Sounder (TOVS) total column water vapor (PW) values, prompted a small intercomparison study between TOVS and other total column water vapor measuring devices. The intercomparisons took place at the Department of Energy (DOE), Southern Great Plains, ARM Cart, Central Facility site (DOE, 1996) during June and July 1996, and the EROS Data Center (EDC) site (43o 44’ N, 96o 37’ W, 485 m elevation) during June 1999. We intercompared the TOVS data with that from an upward viewing microwave radiometer (WMR), sun photometer, as well as available radiosonde (sonde) and NOAA experimental sounding data, since all determine total column water vapor (PW). These devices make their measurements on differing spatial and temporal scales, implying that they might not detect the same amount of PW in a given atmospheric column. The likely influential factors causing differences in PW values include, the inherent error within each measuring device, natural variation in PW, averaging error (due to measurement frequency differences, for example, 2 minute values for the MWR, one degree by one degree averages for TOVS, and approximately 30-50 minute values for the sonde), mismatched viewing of the atmospheric column both spatially and temporally, maximum height of the radiosonde, and cloud presence. The TOVS and sonde PW values for our period deviate by about +/- 5% on the average, with a range of +/- 0.4 to approximately 69%. Similarly for the MWR and TOVS PW data, although the agreement was slightly better compared to that for the sonde and TOVS data. The mean percent difference between the Microtops II and the TOVS PW values is about 38%, with the actual percent differences ranging between 3 and 54% during this period. These differences are affected by averaging errors, spatial and temporal offset errors, and the presence of clouds. The mean percent difference between the Microtops II and the NOAA experimental sounding PW values is about 5%, with the actual percent differences ranging between 0.8 and 7% during the cloud free periods for our study.