WMOD00477 - Laserfiche WebLink

Home Browse Search

MARCH 1987, MARK HEGGLI, ROBERT M. RAUBER AND 1. B. SNIDER 207 n three occasions during calibrations. Procedures asso- ciated with analysis of data for each of these modes and results of the comparisons are the subject of the following discussion. 1) VERTICAL MODE The largest portion of the 3-week experiment was reserved for data collection in the vertical mode. This dataset included both cloudy and clear weather events. Due to differences in the operating programs of the two radiometers, data were recorded at different rates. The NOAA radiometer recorded data each minute, the USBR radiometer every 2 min. The NOAA I-min data points were averaged and compared to the 2-min values obtained by the USBR radiometer. The total 2-min averaging periods for the two radiometers were simul- taneous. Figures 2 and 3 show the comparisons of the mea- surements of integrated water vapor (centimeters) and liquid (millimeters) for the vertical mode dataset. The vapor data are stratified into three groups:' pairs ob- tained when liquid water was recorded (liquid water ~ 0.05 mm), pairs obtained when little or no liquid water was recorded (liquid water < 0.05 mm) and all data points. Statistics are given in Table 1. A high correlation (0.95) was obtained for the mea- surement of vapor in the total dataset. However, the correlation was lower (0.79) in periods when liquid was present. The mean absolute difference and the rms dif- ference for no liquid and all points were between 0.05 1.00 0.75 E .!J 0:: 0 0... <( > cr: 0.50 W I- <{ :3 0:: OJ .. , IJ) ::J 025 0.08.00 0.25 0.50 NOM WATER VAPOR (em) and 0.07 cm. The values were higher (0.07 and 0.08, respectively) in the presence ofliquid. These differences compare favorably with a rms difference of 0.08 cm for two collocated radiometers reported by Hogg et aI., (1983). Measurements of integrated liquid water were con- sidered when either radiometer reported a threshold value of 0.05 mm. For this comparison, a correlation coefficient of 0.99 was obtained with an average ab- solute difference of 0.02 mm. The rms difference was also only 0.02 mm. 2) SCANNING MODE Direct comparison of the radiometric measurements during azimuth scans offers two advantages over ver- tical comparisons. The primary advantage is that the integration path extends over a larger atmospheric path length, resulting in larger values of the measured pa- rameters. This is particularly helpful when only low concentrations ofliquid water are present. The second major advantage is that a wide range of values can be obtained, particularly in the liquid measurement, due to the large temporal and spatial variability of liquid water in an orographic environment. This variability, however, requires that the antenna beams be closely coordinated in space and time to make comparisons valid. Six coordinated scans were made during a 3-h storm period on 28 November 1984. The scans were done at a 150 elevation angle between the azimuths of 1730 &,. 00 .... %~O ~o. ," (DO; is~ lo<f. Q~'go :0 o"tl8...... 0 og8;~.~. 0 00 OCl(l) .u~... oo~o 0 o+;ffi ..,~ o~o o -~ ,," 0 .. "0 ,,'0 o~ 0 :.... ..:~ ' ...... !+ o~ -: .p .0, !ii\fr; . 0.B~3d5 ~cirt:=S~lO _._...Y.... P8:f.ei~t~~~rr~~t > = '~~<j'P' Average a'R'Mg1,,= d1l8~~ . .669 Number sy~t&'; : T7 All ~;rs Corielale:n1 cae ,'(tent '" 0.948 Average a'R'Mg1,,= .~7src1rce . .054 Nu~~~r Po" end ,811 ~~t~;~~~22 .... y.. ~\f,' . 0.8704 ~rr:=~~r 0.75 1.00 FIG. 2. Scatter plot of integrated water vapor retrieved by the NOAA and USBR radiometers . from data taken at Steamboat Springs, Colorado. Statistics are provided for vapor retrieved in the absence ofliquid water (0), in liquid water (+), and all points.