SR SSD 2002-14

Technical Attachment


Gary J. Maniscalco, NOAA/NWS Forecast Office, Mobile, Alabama


This paper compares and contrasts two cases of cold season dense fog development during the evenings of January 12 and 14, 2002. Each event occurred in the wake of widespread rainfall. Regional rainfall distribution, surface and sounding data are examined in order to analyze the conditions prior to dense fog development. Surface visibilities fell below 1/2 mi each evening at both Mobile and Pensacola. This critical visibility value was not forecast in the 1800 UTC TAF for either station. Further, the visibility was not accurately forecast by the 1200 UTC NGM or AVN MOS guidance. The goal of this paper is to investigate reasons why both MOS and WFO forecasters failed to accurately forecast the visibility for each case.


A surface trough (not shown) moved through the region between 1200 and 0000 UTC each event day. Within the extratropical cyclone warm sector, widespread rains occurred over a large portion of the central Gulf Coast (Figs. 1 and 2) prior to dense fog formation. Although not shown, .54 micron visible satellite imagery revealed a sharp clearing line behind the cold front just prior to sunset. In both cases, the line of clouds moved east of Pensacola by 0600 UTC. Surface high pressure was found to be building eastward into the region (Figs. 3 and 4).

Upon examination of the 0000 UTC, Slidell, Louisiana sounding for both cases (Figs. 5a-b), cold air advection was present within the boundary layer on January 15, while the January 13 case was more characteristically neutral thermal advection. For comparison, these soundings are similar to prototypical Southeast US cold-season dense fog soundings discussed by Croft, et al., (1997). In these cases, dense radiation fog occurred under weak, backing (cold air advection) and/or neutral advection patterns along with deep tropospheric dry layer air ( > 19 C dewpoint depression) above. As in Croft, et al., (1997), each of the days in this study was characterized by a deep layer of dry air (300-925 mb in Fig. 5a and 400-950 mb in Fig. 5b) along with a pre-existing near-surface frontal inversion. It should be noted that cirrus clouds on January 15 advected east of the area after 0600 UTC. In the latter case, it was readily apparent that radiative cooling would later take place during the overnight hours and act to form a boundary layer inversion.

Overview of Event SIL 00 UTC Soundings (Y-yes, N-no)


13 Jan 2002

15 Jan 2002

Backing of low level flow w/height?



Presence of deep tropospheric dry layer?



Cirrus Clouds?



Presence of initial low level inversion?



Table 1 - Overview of pre-existing conditions obtained from Slidell, LA Soundings before the onset of cold season dense radiation fog.

Due to a relatively colder airmass that was slowly advecting over warm and moist ground, fog formation was hypothesized to be enhanced upon condensation each night. The NGM and AVN MOS guidance both forecast visibilities to lower between 0000-1200 UTC (Figs. 6a-d). As is shown in the figure, visibilities lowered well below the MOS forecasts for each event. For reference, the following are surface visibility categories (statute miles) for NGM and AVN MOS:



















Of the two event days, the NGM MOS forecast the driest conditions on January 15. Average surface dewpoints were ~ 5 F too dry at Mobile and ~ 3 F too dry at Pensacola. Observed nighttime surface wind speeds were also lighter (1-5 kt on average) for both days and at both sites. Similar to the NGM MOS, the AVN MOS was drier in the January 15 case. Average surface dewpoints were ~ 7 F too dry at Mobile and ~ 7 F too dry at Pensacola. However, unlike the NGM MOS, the AVN MOS accurately predicted the decrease in surface wind speeds on January 13 for both sites. However, in the January 15 case, the AVN MOS overforecast the surface wind (~ 4-5 kt on average) for both sites.

Figures 7a-b show the 1800 UTC TAF verification. The observed surface visibility has been converted into the AVN MOS categories (shown above) so that direct comparisons could be made. In general, surface visibilities fell below Category-3 at Mobile shortly after 0100 UTC in both cases. For Pensacola, the onset of dense fog was generally after 0500 UTC in both cases. Dense fog ended in both events as the surface high further developed into the region, which allowed for increased cold-air and dry advection. As a side note, the 1800 UTC TAF forecaster anticipated both the building surface high to produce slightly higher surface wind speeds (7-10 kt) and a relatively weaker near-ground radiation inversion much earlier in the event. However, this did not occur. As a result, the observed decrease in wind speed was hypothesized to have contributed to large errors in the surface visibility.


From Meyer, et al. (1997), the MOS approach to the prediction of visibility correlates predictand data (local weather observations) with combinations of predictor data (output from dynamical models, surface observations, and geoclimatic information). In applying MOS to the prediction of visibility and obstruction to vision, the visibility and the obstruction to vision were treated as categorized predictands. The MOS equation regression process was allowed to continue until a maximum of 18 predictors were chosen or until none of the remaining predictors contributed an additional 0.1% to the reduction of variance for any one of the predictands. For most regions and projections, 18 predictors were chosen. In the 6- through 12-h primary equations, the obstruction to vision observation or a relative humidity term were often chosen as the first predictor. At later projections, the dew-point depression, the 10-m wind speed, and the relative frequency of visibility less than 3 mi contributed most of the reduction of variance. Additional frequently chosen predictors included the high ceiling height/visibility category derived from observed predictors, wind components at levels below 700 mb, 950-mb relative humidity, and precipitable water. Time-averaged predictors were often chosen for the forecast projections of 9, 15, 21, 27, and 33 hours.

The MOS guidance uses observations at a specific hour as predictors. It does not, however, use a time series of observations (like an antecedent precipitation index) as a predictor. It is quite possible that MOS under-predicted the dewpoint, because there is nothing in the system to indicate that the ground is wet. Therefore, it is hypothesized the damp ground contributed to most of the error in visibility forecast by MOS. With no known research available, these two cases could be considered a "case-dependent systematic error in the MOS."

Given the missed forecasts by both the NGM and AVN MOS and the forecaster, a couple of questions loom. First, would dense fog have developed in the absence of the rainfall? Climatologically, one would expect the answer to be no. Local climatology (Croft, et al.1995, and Garmon, et al.1996) for Mobile shows that only 2% of observed cold-season dense fog cases occurred with surface wind flow >7kt. No such climatology has yet been developed for Pensacola. Further, only 24% of the observed cold-season dense fog events occurred during a northwesterly surface flow regime.

Would nocturnal fog develop in the wake of a previous rain event? If so, would it become dense (< 1/4 mi)? If high level clouds were forecast to remain, longwave radiation would have been trapped and less diabatic near-ground cooling would have occurred. This would in turn have limited the potential formation, extent, and duration of dense fog. However, if no clouds were present, one can refer to local climatology (Croft, et al.1995 and Garmon, et al.1996) for Mobile which shows that under a general variable flow <5 kt, significant fog developed in 76-98% of cases.


Forecasts of dense fog occurrence, its extent and duration are complex considering mesoscale and microscale variability. It is important to consider the state of the boundary layer and how it is expected to change with time. Another important consideration in determining the extent of fog coverage is the degree of boundary layer mixing.

The examination of these two cases indicated that skies cleared, wind speeds became light and a deep layer of dry air existed above a moist ground from previous rainfall events. Thus, for a short time period, dense fog formation was hypothesized to be enhanced upon condensation each night. This is also supported by local climatology for Mobile which showed a much higher occurrence of dense fog for light wind (<5 kt) flow regimes than for slightly higher wind (> 7 kt) flow regimes. For visibility, the forecaster can improve over the MOS by using knowledge of recent conditions (antecedent rain). Further, a thorough knowledge of climatology, implementing local techniques, and a complete understanding of the evolving synoptic situation are each essential in much better fog forecasting.


The writer wishes to express thanks to the following individuals at WFO Mobile: Jeffrey M. Medlin (SOO), Jeff Garmon and Don Faulkner (FICs) and Randall McKee (MIC). Each provided helpful input into this paper. Thanks to Dr. Bernard Meisner (Scientific Services Division, NWS Southern Region Headquarters) for furnishing MOS data and to Paul Dallavalle (Meteorological Development Laboratory) for helpful insight into the MOS.


Croft, P. J., D. Darbe, and J. Garmon, 1995: Forecasting significant fog in southern Alabama. Natl. Wea. Dig., 19, 10-16.

Croft, P.J., R.L.Pfost, J.M. Medlin, and A. J. Johnson, 1997: Fog Forecasting for the Southern Region: A Conceptual Model Approach. Wea. Forecasting., Vol.12, No. 3, pp 545-556.

Garmon, J. F., D. L. Darbe, and P. J. Croft, 1996: Forecasting significant fog on the Alabama coast: Impact climatology and forecast checklist development. NWS Tech. Memo. NWS SR-176, 16 pp. [Available from Scientific Services Division, Southern Region NWS, 819 Taylor Street, Room 10A23, Fort Worth, TX 76102.]

Meyer, F. G., V. J. Dagostaro, and D. T. Miller, 1997: NGM-based MOS visibility and obstruction to vision guidance for the contiguous U.S. NWS Technical Procedures Bulletin No. 431, National Oceanic and Atmospheric Administration, U.S. Department of Commerce.