2. Methods Cloud-top data was obtained from the Geostationary Operational Environmental Satellite (GOES) imager for the day-night cycles of 19, 20, and 21 January 2005. Both visible and infrared (IR) images were used in this study. The Man Computer Interactive Data Analysis System (McIDAS) was employed to view and analyze the remotely sensed data for the three January 2005 dates. A supervised classification protocol was developed to identify cloud-free (CF) pixels near the site of Fresno, California at one-hour intervals during the presence of dense fog and/or near-surface stratus clouds in the Central Valley. The area (km2) of clearing near the urban region of Fresno was estimates using counts of pixels with brightness signatures identified as CF.
Hourly meteorological data from surface observing stations at Fresno (urban site) and Parlier, California (non-urban site) were obtained from the California Irrigation Management Information System (CIMIS). Additional data detailing the amount of various pollutants in the air at Fresno were obtained from the California Air Resources Board. These meteorological and pollution concentration data were examined in concert with UCI parameters to assess the possible influence of any or all of these variables on UCI development.
3. Preliminary Findings The UCI at Fresno, California was first detectable in GOES imagery at 1400PST on all three January 2005 dates. The area of the UCI was observed to be quite variable from hour-to-hour. On 19 January 2005 the first measurable UCI was approximately 66km2 in total area. The 1500PST UCI grew to 371km2 and the final detectable UCI on this date was 519km2. The UCI began to fill at 1700PST and the GOES image for the evening hours became contaminated with higher cirrus-type clouds that made detection and measurement difficult. The 20 January 2005 UCI grew from 44km2 at 1400PST to 171km2 at 1500PST and 198km2 at 1600PST. The GOES Nighttime Fog Product (NFP) (Underwood et al. 2004) was used to detect the UCI after sunset and the UCI was completely filled by 1700PST. On 21 January 2005 the UCI at 1400PST was approximately 48km2 in total area. The 1500PST area estimate was 341km2 and the 1600PST estimate was 329km2. The GOES-NFP image revealed that the UCI began filling before 1700PST (160km2) and was completely filled by 1800PST.
A correlation analysis was performed to assess the linear relationship of UCI development and the variability of both meteorological and pollutant observations at Fresno during the three day period. Statistically significant relationships were found between: UCI area and Solar Radiation; UCI area and Air Temperature; UCI area and moisture parameters including Vapor Pressure, Relative Humidity, and Dew Point Temperature; and UCI area and ozone (O3) concentration.
Finally the meteorological data from Fresno was compared with data from the non-urban observation station at Parlier, California. Distinct differences were observed between the following rural and urban parameters: Soil Temperature (approximately 0.5 to 1.0 „aC greater at urban site prior to UCI detection); Relative Humidity (approximately 4 to 7% less at urban site prior to UCI detection), Incoming Solar Radiation (approximately 25 to 35 Wm2 more SWR at urban site one hour prior to UCI detection).
4. Conclusions This work is a first step in understanding the complexities of UCI development in radiation fog and near-surface stratus clouds in the Central Valley of California. The simple analyses performed are meant to provide direction for further research. The correlations of surface meteorological observations and UCI development suggest that the urban environment is more moisture starved than surrounding rural areas and the air temperature is higher. The fog or stratus deck also seems more transparent to solar radiation (a thinner cloud layer) at the urban site compared to the non-urban site. The concentration of pollutants, such as ozone, suggests that air quality degradation in the urban environment may also play a role in the development of UCI's. This is of course an incomplete look at the conditions accompanying the UCI, as boundary layer data is not readily available for sites in the Central Valley. In the future it is hoped that sounding data can be collected during a number of UCI episodes to provide insight into the conditions aloft during this unique cloud dissipation process.
5. References Lee, T.F., (1987) Urban clear islands in California Central Valley fog. Monthly Weather Review, 115, 1794-1796.
Sachweh, M. and Koepke, P. (1995) Radiation fog and urban climate. Geophysical Research Letters, 22, 9, 1073-1076.
Sachweh, M. and Koepke, P. (1997) Fog dynamics in an urbanized area. Theoretical and Applied Climatology, 58, 87-93.
Underwood, S.J., Ellrod, G.P., and Kuhnert, A.L. (2004) A multiple case analysis of nocturnal radiation fog development in the Central Valley of California utilizing the GOES nighttime fog product. Journal of Applied Meteorology, 43, 297-311.
Figure Captions Figure 1. GOES visible image of Central California. The locations of Fresno and Parlier, CA are marked. The UCI is clearly visible in this 1500PST image from 21 January 2005.