- A post on this blog will occur when severe or disruptive weather is likely in the Chicago area. An inch of snow in January or a non-severe thunderstorm in July would not justify a post in most cases. There will be stretches of several weeks and months with no activity, and then during a severe event there might be several posts on one day.
- I hope to provide a plain-English, common-sense explanation of the weather in the Chicago area, especially when the weather turns severe or disruptive. Occasionally I will try to offer up an opinion on other areas or topics as well. The FAQs (see below) will also hopefully answer some questions and myths that I often hear.
What this blog cannot do
- I cannot keep up with all storms, and severe weather alerts should be obtained elsewhere. Use the NWS website/Twitter or a NOAA weather radio. As a last resort, broadcast media should also be fairly timely for NWS-issued severe weather watches and warnings.
- I cannot provide a daily forecast. If you need to know what to wear, go to the NWS (here and here) or the WGN/Trib/Skilling team here. Those resources are constantly updated by excellent meteorologists who have the best information and tools.
General Weather FAQs -- The Manifesto
- Q: Where should I get my weather?
- A: The best source for forecasts and weather information is the National Weather Service. Go to weather.gov and type in your zip code. In general, you're getting the best information and the most reliable forecast from with the definitive source for severe weather warnings and alerts. In most large cities you'll also get a local NWS office with good meteorologists who have valuable localized experience and can add context, details, and multimedia explanations of significant events.
- Other sources are also useful, but there's really no sense to stray beyond NWS for most people. Weatherunderground has great blogs and other resources. Accuweather produces decent forecasts across the country. WeatherBug is fantastic for very localized information, and the forecasts are OK. CoCoRahs is a cooperative of observers that provides generally reliable local data that is often beyond the WeatherBug network. There are also a bunch of great resources on the Internet elsewhere if you're interested -- email me for a list of sites and Twitter feeds. Locally in Chicago, Tom Skilling and Mike Hamernik are exceptionally good. They and their team post frequently to chicagoweathercenter.com and Twitter, Facebook, etc.
- The Weather Channel and weather.com are the worst option -- the forecasts are intentionally skewed and/or dumbed down, and the content is geared more toward entertainment and ratings rather than accuracy and informational value. (WSI -- the non-Weather Channel assets from the Weather Company, were recently sold to IBM, and it will be interesting to see how Watson and other IBM resources are employed there.)
- Apparently people don't think very much about the source of their weather information, at least if this small survey is to be believed. Despite checking their weather source almost every day, most people reverted to the anonymous default on their phone. Nearly as many people relied on local TV/news, the quality of which varies dramatically. If true, that is certainly a large factor in the poor relationship between forecasts and readers/viewers.
- Q: Why are most weather forecasts in the media so bad?
- A: Great question! Unfortunately there is not a great answer. The truth is that meteorology has made huge strides in the past few years and decades, and the accuracy and usefulness of weather forecasts is quite high. But most people make awful mistakes when interpreting weather forecasts or just don't know how to think about them. Here are some things to remember when you're looking at a weather forecast.
- Errors in the forecast are unavoidable. Most forecasts involve probabilities, as they should -- forecasting anything is an exercise in uncertainty. Unfortunately most people don't think -- or don't know how to think -- about probabilities. It's either "supposed to rain" or not, which is an absurd way to look at the situation
- A forecast should use probabilities, but they are often misunderstood. If a forecast says there is a 70% chance of rain on Tuesday, what does that actually mean? Most readers/viewers who are using that forecast would struggle to accurate translate that forecast to its intended meaning. A 70% chance of rain means that, for the time period and the area in question, a person has a 7-in-10 chance of rain falling on his head. That means several things that most people ignore:
- At 70% it is somewhat likely, but far from a certainty, that rain will occur somewhere within the forecast area. It does not mean that it is 70% likely to rain all day or that 70% of the area will get rain, which are two common misconceptions.
- We should expect no rain on 30 days out of 100 with an identical atmospheric set-up. Most people would see this 70% chance of rain and say "it is supposed to rain" or complain about a bad forecast if it does not rain at a specific location, both of which are silly.
- Getting no rain does not mean it was a bad forecast. Evaluate the correctness of the probabilities with regard to the process and outcome, and don't fall prey to the "it was supposed to rain" nonsense we all hear so often.
- Even if it didn't rain on you, the forecast may have verified elsewhere. A forecast calling for a 100% chance of rain may verify if a storm rolls through far southern Olympia Fields even while it is dry as a bone in Palatine -- they're both in Cook County, but they're 50 miles (and often a world) apart. This is especially true of severe watches and warnings. I often hear complaints about "bad" forecasts when severe thunderstorm or tornado warnings are issued -- those watches/warnings are very often warranted, but if the storms affect less than 100% of the forecast area the people who missed the storm think it was a blown forecast when the opposite is true.
- Forecasts often cover a large area over many hours. To reiterate the prior point, even just one day across the Chicago area entails a massively complex interaction of atmospherics, time, and varied geography. The forecast area for most Chicago weather forecasts and data encompasses more than 10,000 square miles and nearly 10 million people. Conditions vary dramatically across that area. Even smaller areas like Cook County or the city of Chicago are still enormous and can have vastly different weather conditions on any given day.
- It is pointless to provide a single-temperature forecast. Just as all forecasts should include probabilities, they should also include ranges. Not only will a single-point forecast be more frequently wrong, it's not helpful. Nobody cares if it's going to be exactly 66 or 68 degrees tomorrow, they just want to know if it's going to be relatively pleasant -- an afternoon forecast of mid-to-high 60s would be both helpful and more likely to be accurate. I really don't know why forecasts aren't broken out in 5-degree increments for early morning, afternoon, and evening, since those are the time periods people care about when planning their lives.
- Most broadcast meteorologists are not very good. Some are not even real meteorologists, but even the ones who are tend to worry more about ratings than anything else. There are notable exceptions -- Tom Skilling in Chicago is one -- but most network or cable meteorologists are not worth listening to.
- Pictures/diagrams in forecasts are often misleading. Many forecasts use a 5/7/10 day outlook with little pictures of sun, clouds, rain, snow, etc. I think the diagrams are more misleading than helpful. Imagine a weekly outlook with a little snow diagram on four consecutive days -- you're likely to hear a comment on the street that "it's going to snow all week" when the reality may be (and is more likely to be) that each of those four days may see some snow during a portion of each of those days, but we have no idea how likely or how heavy the snowfall may be. The same applies to clouds, rain, etc.
- Forecasts beyond a few days are not helpful for their specifics. If you're trying to plan an event, don't bother looking for specifics more than 3-5 days in advance. Most media services provide 7- or even 10-day forecasts, but the actual numbers on days 5 and beyond are not very good. The better models are quite good at picking up large scale anomalies 7-10 days in advances, but that mostly applies to major cold snaps, heat waves, large-scale snowstorms, etc. Beyond 3-5 days you're not going to get much useful information about specific weather events (temperatures and precipitation). And the smaller details -- e.g., the timing of a rain or snowstorm -- are hard to pinpoint at all, but especially more than a day or two in advance.
- The temperatures in a forecast for a given day are shown across two calendar days. For example, if you're looking at the forecast for Saturday, the National Weather Service (and most local services) will show a single temperature for the high and a single temperature for the low. That high temperature is assumed to occur on the calendar day of the forecast (here, Saturday) while the low temperature is assumed to occur the next morning (here, Sunday).
- For climatological purposes, daily highs and lows are defined as the max and min temps that occur during a given midnight-to-midnight period, regardless of the time of day those high/low temps occur.
- Q: What is the difference between a watch and a warning?
- A: A watch means conditions are favorable for the event to occur. A warning means the event is occurring or imminent. For example, a tornado watch means that conditions are favorable for a tornado to occur somewhere in the watch area. A tornado warning means a tornado is already occurring or is imminent in the warned area.
- A watch means that you should pay attention to future conditions; it is meant to serve as advance notice that trouble could be on the way. A watch does not always advance to a warning but it often does.
- A warning means that the severe event in question is happening right now or is about to happen. Remember that a warning may be valid
- Q: What are the criteria for severe weather?
- A: Here are the levels for various severe weather events
- A severe thunderstorm is producing wind gusts over 58 mph and/or hail at least 1" in diameter and/or a tornado. Lightning, rainfall, and other criteria are not factors in a severe thunderstorm watch/warning, although they may be noted in warning or cited in a special weather statement.
- A blizzard is often used colloquially to mean a severe snowstorm, but the actual definition of a blizzard does not require any falling snow -- the defining characteristics are high winds and low visibility over an extended period. Specifically, a blizzard requires sustained winds over 35 mph and falling or blowing snow causing visibility below 1/4 of a mile for at least three consecutive hours.
- Winter storms vary in their criteria depending on the location and the time of year. Winter storm watches/warnings and winter weather advisories are all defined by state or by NWS local office.
- Q: What is "normal" weather?
- A: Normal values in meteorology are calculated as the rolling 30-year average for a given date. Currently the period in use is 1980 to 2010. For example, the average high and low temps in Chicago on April 12th are 58 and 38 degrees, respectively. Those are just the average of all high and low temps on April 12th from 1980 through 2010.
- Q: Where are official records kept?
- A: O'Hare Airport is the official site for Chicago, and the biggest airport is often the official location for other cities as well.
- The official location in Chicago has changed repeatedly over the years, often which significant implications. For the first 70+ years the records were taken downtown, near the lake. That obviously has implications for temperatures, snowfall, etc. In 1942 the site moved to Midway, which is more urban -- surrounded by a fairly dense city environment and far enough away from the lake to be spared some or all of its effects. In 1980, the site moved to O'Hare, which is surrounded by a far less urban, much cooler environment than that around Midway.
- O'Hare: January 1980 - present
- Midway: July 1942 - January 1980
- Univ. of Chicago: January 1926 - June 1942
- Various locations in Loop: 1870 - January 1926
- Accordingly, when you see a record from before 1980, it may not be all that comparable to what's been recorded at O'Hare since then. Many of the daily and all-time records would be quite a bit different had they been taken at one of the other locations.
- Q: Can it be too cold to snow?
- A: No. It has snowed in Antarctica at ridiculously cold temperatures (far colder than anything ever recorded in Chicago). It is true that very cold air holds less moisture than less cold air, however, and that likely gave rise to this myth. The coldest days in Chicago and other mid-latitude locations are often under Arctic high pressure systems, which are by definition dry.
- Q: What is wind chill?
- A: Wind chill is a calculation designed to capture the additional heat loss a human being would experience due to accelerated heat loss on bare skin due to the wind. If that sounds somewhat arbitrary, it is. There is no perfect way to quantify how wind will accelerate heat loss on exposed skin because everyone's skin is different (fatter, hairier, etc.). And of course most sane people will be wearing heavy clothing when it's really cold, so the effects of wind chill are heavily impacted by how a person is dressed.
- The entire calculation of wind chill was rewritten in 2001, and the change was dramatic. Under the old formula, a stock Chicago cold snap of 0 degrees with a 20 mph wind produced a wind chill of -39 degrees; now the formula says that's only -22 degrees. So take wind chill with a giant grain of salt.
- Sidenote: Don't be the person who uses wind chill and temperature interchangeably. They're not the same thing. Temperature is an objective measurement of molecular movement, while wind chill is a subjective (and varying) guess at how a person "feels" given a certain combination of temperature and wind .
- On that note, when you hear someone say "It was 30 below" or "It was 40 below" in Chicago, that's not really true. It's never been that cold (the all-time record is -27). It would only be accurate to say "the wind chill was [x]" or "the temperature was [x] but it felt more like 40 below with the wind."
- Q: Could a tornado hit downtown Chicago (or another large city)?
- A: Yes, a tornado could hit downtown Chicago or anywhere else. It has happened before in Chicago and it will happen again, given enough time. There is plenty of crappy psuedo-science and faux journalism, along with a healthy dose of availability and recency bias, that has lead to the belief that big cities like Chicago are immune. Tornadoes, especially strong ones, are the products of massive forces that would not be significantly impeded by tall buildings, the warmth of the city, the air over the lake, etc. (An average tornado would make the Sears Tower look small by comparison, and the atmospheric currents creating that tornado are enormous.) Very small/weak tornadoes may be affected by urban environments, but those are less of a threat in any case. It has been several decades, but strong, destructive tornadoes have hit the city of Chicago on several occasions.
- Most recently, in 2000 a (thankfully moderate) EF-3 tornado went through downtown Fort Worth, Texas. The devastation was still significant. The DFW area is among the most vulnerable to a devastating tornado in a dense population.
- In 1990, an F-5 tornado (the most destructive on the Fujita scale) hit Plainfield, just 35 miles from the Loop. That storm was isolated but enormously powerful.
- In 1967, a deadly tornado outbreak included one in Oak Lawn that crossed the Dan Ryan before going out over Lake Michigan near the 79th Street beach.
- In 1876, a very destructive tornado went right through downtown before moving out over Lake Michigan, reportedly with multiple vortices.
- Most modern high-rise buildings are built to withstand tornadic winds, although (thankfully) they are rarely if ever put to the test. Tornadoes have recently hit high-rise buildings in Fort Worth, TX, for example, resulting in extensive damage but no structural failure. The biggest risk is broken and flying glass, which presents a lethal danger to both building occupants and pedestrians. Take shelter in the designated shelter on your floor or else the inner-most room with no windows (hall, bathroom, stairwell, etc.).
- Q: Do tornadoes favor certain areas, trailer parks, etc.?
- A: No. Tornadoes do develop in certain areas more frequently, but they are fairly small phenomena (anywhere from 50 yards to a quarter mile wide, most commonly, up to a maximum width of a mile or two) when compared against the large area of a city/county/state. Within a given area (say, 50 or 100 square miles) the atmospheric conditions will be roughly equal such that the distribution of tornadoes will be uniform and occurring randomly over time. If tornadoes seems to hit certain structures like trailer parks more frequently, that is because trailer parks are more prevalent in the tornado prone area and/or less secure than other structures there and thus more likely to sustain newsworthy damage. The vast majority of tornadoes are strong enough to flip or destroy a mobile home but not strong enough to do structural damage to a well built home.
- Q: What should I do in a tornado warning, especially if I'm in my car?
- A: A car is a very, very dangerous place to be in a tornado. Tim Samaras -- one of the fathers of storm-chasing and tornado research -- was killed in his car by a tornado. He was much smarter, more informed and more diligent than the rest of us, so if he can be killed that does not bode well for anyone else who takes a chance.
- Immediately seek shelter in an interior, windowless room on the lowest level of a well-built structure. Put as many walls between you and the tornado as possible. Do not attempt to outrun a tornado and do not seek shelter under an overpass (the constricted airflow and raised elevation of an overpass will do more harm than good). As a last resort, if no building is available, do not stay in your car -- lay flat in a ditch or low-lying area and cover your head.
- Do NOT be like this guy, who needed a massive amount of luck to escape with his life.
- The "Windy City" myth
- Chicago's "Windy City" nickname has murky origins but nothing to do with the weather, strictly speaking.
- The fact is that Chicago is very average in terms of average annual wind speed and there are windier places in nearly every part of the country. Chicago ranks 76th out of 300 NWS offices nationally and doesn't even crack the top 10 for major metros (coming in 12th). Much of the Northeast corridor -- including Boston and New York City -- is considerably windier than Chicago.
- Chicago's windiness has taken on a life of its own in most people's minds, though, as perception seems to exceed reality. The location of Chicago's urban canyons right along Lake Michigan does enable significant wind channeling, particularly during memorable events like snowstorms and cold days, and that is probably why people think Chicago is windier than it really is.
- Q: Is global warming real?
- A: Politics aside, climate change is real -- the climate is, by definition, always changing. And the data are very clear that the climate has been warming overall in recent decades. That said, some important thoughts are often overlooked.
- We know very little about how much the climate has warmed -- the reliable record of global satellite data is only a few decades old -- let alone how much or if it will continue to warm. Ice cores and other methods are reasonably reliable but not perfect.
- We know even less about the causes of the warming that has been observed. Some of it is likely natural, and some of it is likely anthropogenic. Where we lie on that spectrum is unclear.
- The record is unequivocal that massive, pre-human climate changes are a fact of life on this planet.
- It also stands to reason that adding significant amounts of greenhouse gases -- the effects of which are very well understood -- would have some effect on the climate. Likewise for land development, since asphalt, concrete, buildings, etc. are much more effective at heating the atmosphere than natural surfaces. Again, the magnitude is uncertain but the direction is very clear.
- Our atmosphere is almost infinitely complex and dynamic. Science has made massive strides in the past few decades in forecasting large-scale and near-term events, but we're still flying blind on many levels. And if we can't accurately model the atmosphere a month or a year in advance how can anyone pretend to pinpoint climate changes down to one degree -- or even fractions of a degree -- 50 or 100 years from now?
- Precision may be foolhardy, but if any significant changes are intolerable, acting to constrain the potential changes would be prudent (an a potentially intractable an exercise in tradeoffs).
- The record would suggest that intermittent or "mini" ice ages are more likely than not. Even a "mini" ice age could well have effects that are less palatable than global warming.
- The planet has wobbled on its axis, and those effects may well trump all others.
- Most of the planet's surface is covered by water, and most of the land mass is uninhabited. So please don't be the person who points to unusual temperatures in one location -- even over a relatively large area like the United States -- as evidence for or against an argument.
No comments:
Post a Comment