Predicting Fog One Morning a Few Weeks Ago, We Were Overlooking A

Predicting Fog is never in the radio broadcasts, but is readily available to us by other means. That is not to say such fine details will One morning a few weeks ago, we were overlooking a be right all the time, but they are at least letting us see what foggy Puget Sound planning a cruise to the San Juans for the best science on the subject has to offer. the next day. For the past few days we had had the same weather pattern—fog at daybreak, burning off sometime This exercise might also serve as an introduction to one of in the morning. But it was already noon and still just half a the very best marine weather sources on the Internet, but mile visibility. The local VHF marine weather forecast was one that is only slowly being discovered by sailors, even no help. Strangely enough it did not mention fog at all. So though it has been right in front of us for about a year or the challenge we address here is how do we predict on our so. The trouble with finding it online is the word “weather” own when such a fog will lift, and what is likely to happen (even “marine weather”) is a complete disaster in any the following day? (We can guess right now that fog forma- Internet search engine. It might be reassuring to know that tion is somehow the opposite of fog dissipating, so if we can the best source is the National Weather Service (NWS) predict one, we can likely predict the other.) themselves, but this does not help us find it. Commercial services, which rarely add to the NWS content, manage to The fog being experienced in this example comes about grab all the weather traffic. So here is a step by step route to when we have a moist air mass over us in a relatively high the NWS data. pressure. High pressure means clear skies, and without cloud cover to keep us warm after the sun goes down (1) Go to this link: http://forecast.weather.gov/zipcity.php. (greenhouse effect), the surface air cools rapidly as the sur- (If for any reason that link stops working, go to www. face heat radiates away at night. A drop of 15 to 20° F is not starpath.com/navpubs for an updated link. There are other uncommon in these conditions. When the air temperature things of interest to navigators there as well.) drops to the dew point, water vapor in the air condenses into fog. It will remain foggy until the temperature rises (2) Enter city, state, or zip of the nearest town to the water- above the dew point. This is called radiation fog. It gener- way of interest. You will then get to the page that is used for ally forms on land and spills down onto the water. land based weather, an example of which is shown in Figure 1. This is the same page you might bookmark for decid- It is a mystery why the marine forecasts did not discuss ing what clothes to wear to work, or whether it will rain on the fog in that example—had they discussed it, however, an event you care about. This is the “secret source” we are we likely would not be discussing it now. It could be the looking for! automated light beam that measures the visibility was not working, or its path happened to be located in a pocket of (3) Now drag and zoom the inset map so that you see clear air. Obviously no one contributing to those reports clearly the waterway you care about. You can actually could see Puget Sound. Normally fog predictions are excep- open this page for New York City, then zoom way out, drag tionally good, even when wind predictions are very wrong. to Florida, and zoom back in to get weather for Key West In a moment we see exactly how anyone could know exactly waters. It does not matter how you get to this page with this when this was going to lift and also when it might form interactive inset map. again. (4) Once you have your waterway in view, double click the Air temperature and dew point are air mass properties that water at that location. It is important that you click water are uniform over large areas and easy to measure and fore- and not land. You should see a small red square show up cast. Wind over a wide mountain valley like Puget Sound, that marks the region of your forecast on the inset map. The on the other hand, can change dramatically with a few page will now have magically switched from land weather to degrees rotation of the isobars. It is not uncommon in this marine weather. You triggered this by clicking water rather area to get a wind forecast of “variable 5 to 15.” Variable than land. Again, you may want to bookmark this location, means the wind direction is unknown; “5 to 15” is another which gives all of the standard marine weather information way to say “0 to 15,” which means the wind speed is also we get various places. The special information is in the next unknown. The same broadcast could then forecast after- step. noon temperatures and dew points accurate to 1° out of 60 or so, and they would most likely be right. So we can predict (5) Now scroll to the bottom of the page to find a link on the behavior of fog as soon as we find access to these accu- the right called: Hourly Weather Graph and click that link. rate predictions for air temperature and dew point. Then uncheck all data boxes except Temperature, Dew Point, and Surface Wind (kt), and click Submit. You will see But I should stress once again, even though we are now fog a display like that of Figure 2. Later you might want to look sleuthing on our own, normally maritime fog or reduced at the detailed sea state forecasts, but for now we concen- visibility prediction is very accurate and dependable when it trate on fog. This is the picture that tells you when the air is reported. What we are about to do just fine tunes reports temperature will fall to the dew point and then when it is that might exist, and also fine tune other observations, predicted to rise above it, marking the formation and dis- such as what hour of the day we expect a front to cross our sipation of fog for that region. waterway with a sudden change in wind. This type of detail To learn more as you study this process, watch the live data very closely from the nearest light house or reporting buoy (www. ndbc.noaa.gov) as you visually watch the range of visibility on the waterway. You will notice correlations between visibility and the difference between the two temperatures, even down to the fraction of a degree in some cases.

Keep in mind that you can actually view this data (hourly graphs and live reports) from a smart phone underway on many coastal and inland waters, just as you can see the weather radar images underway that show precisely where rain might be located and how it is moving.

Wind shifts at a frontal passage are always a sharp veer (shift to the right), so by looking at these same hourly graphs you can predict when a front will cross. If you are racing and expect from this data that a front will cross during your time going to weather, it will pay to be on the right hand side of the course, and so on. To my knowledge, this is the only place we have such detailed information, so readily available.

To practice using this data for wind or fog predictions, check a weather map from the Ocean Prediction Center (www.opc.ncep. noaa.gov) to find a region with a nice High in place to look for fog, or another one with a front about to pass in a day or so. Then go to that region with this weather source to see what you can learn about the timing of fog and wind.

The technology at hand these days is quite remarkable. You can even call a buoy to ask for its present weather and sometimes sea state. Call 888-701-8992 and enter station number 28963 (BUZM3, Buzzards Bay, MA) for a sample. See www.ndbc.noaa.gov/dial. shtml for details and list of station numbers. Not all stations give the dew point.

For completeness it must be said loud and clear that there is another kind of fog, called sea fog, and it does not behave this way at all. The resource we have discussed will not help in the detailed timing of sea fog epi- sodes, but the general forecasting of them should be very good. Sea fog is formed at sea and moves around with the prevailing wind. Radiation fog is limited to light air, sea fog is not. Radiation fog has trouble developing Figure 1. Clicking the interactive map gives land weather for land clicks and marine in winds over 10, because the air is not still weather for water clicks. Then click the bottom right link called Hourly Weather enough to cool down and it mixes with other Graph for timing details. layers. Some 3 to 5 kts of wind is most favor- able for radiation fog formation. It lets the fog form then moves it on to pile up somewhere. Less than 3 kts or so leads to waist deep patchy fog. Dense sea fog, on the other hand, can coexist with 30 kts of wind, just as well as with calm air. It is another topic altogether.

Figure 3. Radiation fog spilling off of Whidbey Island into the Eastern Strait of Juan de Fuca. Aerial photo by the author.

Figure 2. When the air temperature drops to within 1° or so of the dew point we can expect fog to occur. For radiation fog, this occurs most often when the wind is light and the pressure is high. On Sat, we expect fog after about midnight and then clearing again by 9am. This data does not call for any fog the next night. Relative humidity does not help with this prediction. It is more a measure of com- fort than of navigational value. Notice that the light westerly switches to a light easterly within an hour of sunrise. These sudden wind changes are usually more interesting when the wind is stronger, but when racing it matters always. Click anywhere on this graphic plot and you will get a table of the numerical values for each hour. From the starpath marine weather course Notes on Fog Prediction (This is original article. There is some duplication, but a new example.)

We are sitting in Seattle on Thursday, planning a trip to the San Juan Islands on Fri, Sat, and Sun. We have noted that it has been very foggy here each morning, then by 10 am or so, it starts to clear. Strangely enough, however, the local marine forecasts have not mentioned fog. Our task then is how to predict the fog ourselves in these conditions. The fog comes about because we have a moist air mass over us in a relatively high pressure, so it is clear at night. This is radiation fog that has spilled down onto the water.

Go to the local NOAA weather for your area (ie forecast.weather.gov/zipcity.php), then be sure the cursor is in the water and not on the land and look for the reports. Look up our article called “Workhorses and Secret Sources,” which describes the use of this resource in more detail.

Then go down to the Hourly graphs, and select temp and dew point, and maybe wind and waves, and look for the report. Here is the example we did. The key issue is we get fog when air temp = dew point, so looking at Fri am we see that there is likely fog till about 8 or 9 in the morning, then it should be lifting. There is no fixed tolerance on how close do they have to be to make fog, but equal is good enough, and so is with in 2° I would guess. At the moment, for example, it is 11 am on Thursday in Seattle, 30 miles to the south, and the temp is 0.5° above the dew point at a lighthouse we can see, and we estimate the visibility is now about 1 nmi. it was thicker earlier this am, down to about 0.5 miles or less.

Visibility, however, is sensitive to a lot of things, and is never the same in all directions, and it also depends a lot on wind when it is radiation fog. If the wind builds at all, it will be gone. Sea fog, on the other hand, could just get worse with increasing wind, but that is not the situation at hand now. So here is your exercise, whenever you see fog, go to this report for your area and watch the temp difference between the air and DP as you monitor the visibility. That way you can then make your own forecasts.

I think we see differences in visibility with just a few tenths difference change, ie more fog for a difference of 0.1° than with a difference of 0.5°.

—david burch www.starpath.com