WunderBlog Archive » Weather Extremes

Category 6 has moved! See the latest from Dr. Jeff Masters and Bob Henson here.

Which NWS Offices Have the Hardest Job Making Forecasts?

By: Christopher C. Burt, 9:00 PM GMT on May 08, 2013

Which NWS Offices have the Hardest Job Making Forecasts?

A Mr. Cesar Ruelase of Chicago made a bit of a stir last week when he accused WGN TV’s weather man Tom Sullivan of being a “weather guesser” rather than a “meteorologist”, see his web site on the subject. Of course, Tom Sullivan, like most TV weather people, get much of the fundamental data used for their weather forecasts from the NWS (National Weather Service) local offices. Obviously, this was just an attention-grabbing stunt or joke on Mr. Ruelase’s part. But it made me think just which NWS offices in the U.S have the toughest job predicting the weather for their forecast areas? My top candidates for such are herein. I’m really sticking my neck out on this one!

There are 125 different NWS centers in the NOAA/NCDC network spread over 6 different regions. By region they break down as follows:

EASTERN: 25 centers covering Maine, Maryland, Massachusetts, New Jersey, New York, North Carolina, Ohio, Pennsylvania, South Carolina, Vermont, and West Virginia.

SOUTHERN: 32 centers covering Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, New Mexico, Oklahoma, Puerto Rico, Tennessee, and Texas.

CENTRAL: 38 centers covering Colorado, Illinois, Indiana, Iowa, Kansas, Kentucky, Michigan, Minnesota, Missouri, Nebraska, North Dakota, South Dakota, Wisconsin, and Wyoming.

WESTERN: 24 centers covering Arizona, California, Idaho, Montana, Nevada, Oregon, Utah, and Washington.

ALASKA: 3 centers all covering AK alone (Anchorage, Fairbanks, and Juneau).

PACIFIC: 3 centers covering Hawaii, Guam, and American Samoa.

Here is a list of all the specific NWS forecast centers within each state and region that provide the weather forecasts for their respective areas (outlined on the map). Most are city-based but a few have area-wide designations. The complete list of all the NWS sites and NWS organizations can be found on their web site.

Never mind that some might find it odd that Wyoming and Colorado are in the ‘Central Division” or New Mexico in the ‘Southern Division’. Just how which state ended up in what division I’m not sure of. Note that the NWS regions are different from the various NOAA/NCDC climate zones that NOAA/NCDC use for their data collection:

The NWS forecast regions are not the same as the NOAA/NCDC climate regions outlined in the map above.

There are obviously some centers that must deal with specific extreme weather events that are particular to their forecast areas, like Oklahoma City for tornadoes or Miami for hurricanes. However, those folks have other agencies that come to their aid so far as day-to-day forecasts when severe weather threatens; the National Hurricane Center for anywhere when a tropical storm threatens or the National Severe Storms Prediction Center for anywhere that tornadoes or severe thunderstorms are forecast.

But in terms of difficulty making a normal daily forecast (the type that Mr. Ruelase was so worried about) some centers have more of a challenge than others. Those would be the ones that cover large areas that include many different climatic zones and microclimates as typical to mountainous or coastal regions. This problem is compounded if that area also has a dense population with inhabitants occupying all those various zones. So I would assume this makes the centers that cover forecasts for the West Coast, Southwest and Rocky Mountains particularly challenging.

For instance, in the Chicago area it is likely that any given day within their forecast area will most likely not see a temperature range of much more than 20° from the coolest (probably Lake Michigan shoreline) to the warmest (somewhere within the urban heat island or in their most southwestern-most inland sites). Likewise, precipitation or snowfall amounts are not likely to vary too much within the greater Chicago area (aside from when isolated thunderstorms might drop 5” in one spot but leave other nearby places dry). This can be said for most of the forecast areas east of the Rocky Mountains.

However on the West Coast, and California in particular, any typical summer day will see a temperature spread of 40° or 50° within one forecast zone and during significant winter storms precipitation amounts might vary from 1” to 20” in a single zone given the effect of local topography. This obviously makes precision forecasting considerably more difficult than those regions without extreme microclimates.

How can you forecast this? A Pacific storm with a tremendous inflow of low-level moisture swamped portions of the San Francisco Bay Area over a four-day period on March 13-16, 2012. San Jose airport picked up a paltry total of 0.68” but just 20 miles away, in the Santa Cruz Mountains, Scott Creek picked up a flooding 20.32”. Generally in a situation like this the NWS-Monterrey office would forecast a broad brush of precipitation amounts like “one inch in the drier inland valleys to five inches in the wetter coastal mountain locations”. It would sound absurd, however, to say “from less than inch in the driest valleys to more than twenty inches in the wettest mountain locations”. With 7 million people in their forecast district, the brush can only be so broad before it becomes meaningless.

There is another reason the west coast NWS offices have a hard time with their forecasts. The most significant weather events for coastal California, Oregon, and Washington, are when major Pacific storms head toward the coast. Unlike the rest of the country there are no weather stations to their west (out in the Pacific Ocean) to provide valuable METARS data or precipitation totals. So the weather that is headed their way has no detailed surface data helping them forecast what is to come. For the most part, they must rely on satellite imagery and upper air data. Coastal fog is another difficult issue they must deal with. Most summer days experience varying degrees of coastal fog development that greatly affects temperatures depending upon how far the fog penetrates inland, how deep the marine layer becomes, and how quickly the fog burns off over the course of the daytime. This issue is particularly burdensome for the NWS-Monterrey center (which covers the San Francisco Bay Area) and the NWS-Los Angeles and San Diego centers. Of course, all three of the centers also serve huge populations (about 20 million in all, or about 7% of the entire U.S. population) who inhabit almost every square mile of their respective forecast areas. In the Rocky Mountains and inland Southwest, elevation by and large, determines what temperatures might be expected between valley and mountain locations. This is not the case along the West Coast where it is often warmer on the mountaintops than in the valleys because of marine air inversions.

A typical summer afternoon temperature spread in the SF Bay Area: from the 50°s at the coast to the 100°s twenty miles inland. Note the 87° temperature on the peak of Mt. Tamalpais in Marin County (the site inland from Bolinas Bay on the map above) and 72° in Mill Valley at the base of the mountain (the figure directly under the 87°). The two locations are about 2 miles apart with the Mt. Tamalpais site being 1,500’ higher than the Mill Valley site. Low-level marine air is causing it to be hotter in the coastal mountains than it is in the coastal valleys. Again, like forecasting precipitation amounts, it becomes absurd to say “the high temperature today will range between 50° and 100°”. wunderground.com map for a (typical) June afternoon last year.

So, although every NWS forecast area has its local weather challenges that can complicate their forecasts, I would hazard to posit that the NWS-Monterey and NWS-Los Angeles centers have, on average, the greatest difficulty in making specific temperature and precipitation forecasts. Their jobs are also complicated in so far as the large populations they serve. One thing, however, is that extreme life-threatening weather is relatively rare along the West Coast so at least the good folks at those offices normally do not have to worry about that quite as often as their central and eastern counterparts.

Christopher C. Burt
Weather Historian

Extreme Weather forecasting

The views of the author are his/her own and do not necessarily represent the position of The Weather Company or its parent, IBM.

Reader Comments

Fire weather in California! Seems like they deal with variable winds sometimes with fires, too, not just Santa Ana winds.
Very educational !! Thank you
I have noted though that many forecasters call wherever they happen to live
"particularly difficult to forecast for because {insert reason} )

I'll go out on a limb here and say the DC area is NOT particularly difficult to forecast for.  I found the FL panhandle, where I did my graduate work in TLH,

I think the first link at the top of your blog header may have moved (?) or doesn't work, sorry.

Do you mean Tom Skilling? (not Sullivan) Skilling is a WGN main met here in Chicago.
Too bad about this disgruntled guy, Ruelase who got his name in lights for going after Mr. Skilling.
WGN and Mr. Skilling's broadcasts are usually very comprehensive. And, to their credit, Mr. Skilling and his team probably get more air time than any TV mets in the country. Their broadcasts, seminars and print/social media every day contain extensive weather history, experienced modeling interpretation for our area, and are full of weather facts and data. Their/his graphics and communication of weather to the public is state of the art--they get the message out in a modern, relevant, yet scientific way, in my experience.

Thank you for a very interesting blog. Reading blogs here at WU from people who live in the mountainous areas, they seem to write a lot about their micro-climates and all the variables to deal with in a local way. I can see how those mets do have their hands full.

Well I live in San Juan , PR and they try ( The Weather woman or man) but they never get the forecast right because there r so many factors involved when trying! They try really hard I might add! So, they wld be better off just to say it's going to rain or we r having Thurderstorms or it's going to be hot and dry at 30% chance at any given time on the island! That way, they wld have all the bases cover! Yes?
I was once told that in the 1960s - when of course forecasting was far less advanced - RTE in the Irish Republic used to offer three forecasts - ranging from most to less likely. Here in the north west of England I cross a ridge from home to work, a distance of maybe eleven miles. The weather can be strikingly different - in snow, the difference between a problem and no problem, and often between an overcast damp day, and dry with broken cloud. The ridge is less than 1,000 feet high.
I would tend to agree that California's coastal NWS offices have the toughest job. For instance, I've seen summer days where just Los Angeles County alone could have weather ranging from morning frost in the mountains, to 50s and foggy in Long Beach, to 80 and breezy in Malibu, to 90 with severe thunderstorms in Palmdale, to 105 and sunny in Pomona. And, as you noted, the seasons can really add a lot of variety to the mix.

(The easiest would have to be South Florida. Unless there's a tropical cyclone threatening, the summer forecast is the same for days and weeks on end: mostly sunny, hot, and humid, with highs in the 90s, lows in the upper 70s, and a high chance of afternoon/evening thunderstorms. And in the winter, sunny and dry, with highs in the low 70s and lows in the high 50s.)
weatherhistorian has created a new entry.
Murphysboro , IL not Indiana was where 239 people lost there lives during the infamous Tri State Tornado in 1925
I find that my small state of Connecticut is very difficult to forecast only because we are governed by THREE NWS centers! One Albany NWS predicts the weather for Litchfield County. The Boston NWS makes their forecasts for Hartford, Tolland and Windham counties, and the NWS out Out of Upton NY predicts theirs for New Haven and New London counties. I am a meteorologist living and predicting my forecasts for Torrington in Litchfield County on Facebook. To me this is totally absurd. The timing and warnings come out at different times from each of the three NWS offices for this small state.The official reporting station representing Connecticut comes out of BDL ,Bradley International Airport in northern Hartford County. You can imagine the confusion from viewers especially in the winter when two counties have winter storm warnings issued and the other four counties have nothing issued till later. My solution is to have each county have one main reporting station and the totals divided by 6 to give us one average of precipitation, temperature and other weather data. Bradley Airport should not be the one station that provides the historical averages for this state. AND one NWS should govern the entire state for the forecasts.