• Weatherwatch
• Virginia's Special Case
• Why "Greenhouse Warming" is Winter Warming"
• Forecast Verifications
• NWS Long-Range Forecasts

  ---

  Advisory Staff

  Dr. Patrick J. Michaels, State Climatologist

  Philip J. Stenger, Research Coordinator and Advisory Editor

  Paul C. Knappenberger, Senior Analyst

  Peter D. Schwartzman, Dustin Hux and Stephen Gawtry, Research Assistants

  Beth Leverich and Tracy Lewis, Undergraduate Assistants

  ---

  
  

  Weatherwatch:

  Year 1996: Temperature Departure -0.56 F, Precipitation 54.69 inches (129% of normal).

  Year 1997 through February: Temperature departure +3.30 F, Precipitation 5.07 inches (80% of normal).

  
  
  December 1996
  • Statewide temperatures for December 1996 averaged 2.81 F above the long-term average-still within the range of normal variability.
  • Precipitation totals for December averaged about 1.72 inches above normal statewide at 4.86 inches. This was outside the range of normal variability, at about the 80th percentile.
  
  
  January 1997
  • January 1997 temperatures averaged across the state were about 109 F above normal.
  • January 1997 statewide total precipitation averaged 0.62 inches below the long-term mean at 2.65 inches.
  
  
  February 1997
  • The statewide temperature average for February 1997 was well above normal by about 5.5 F.
  • Precipitation totals averaged 2.42 inches for February 1997, slightly below the long-term average.
  
  
  Late Moisture Indices
  
  • Palmer Drought Severity Index values for the week ending May 31, 1997 indicated that deep soil moisture around the state was near or slightly above normal.
  • Crop Moisture Index values for the same period indicated that near-surface soil moisture conditions across the state were running fairly close to normal.
  
  

  A Note on the Data

  Weatherwatch maps are based upon a variety of different data sources. Preliminary reports from the dense network of the National Climatic Data Center are used for the first two months, while a much reduced set is used for the third month, consisting of selected National Weather Service stations, a few Climatic Data Center Observers, and (for precipitation only) some volunteer data where important anomalies are suspected.

  Please contact this Office before using these results for legal or economic purposes.
  

Cyclones

SPIN DOCTORS

It's been several years since we've given "good-ol'-low-pressure-systems" a run through the Advisory mill, and the new awareness that climate doesn't always stay the same provides as good an excuse as any to take up that subject again.

It has come to our attention that most people think that a low pressure system is just some kind of mysterious vertical hole in the atmosphere that creates a need for TV weather forecasters. Far from it-they're actually a little more complicated than both. In general they are far from being perfectly vertical and they often don't do what they're forecast to. At any rate, here's our stab at the garden-variety low pressure systems that dot the weather map every evening. We'll even mange to conflate these windbags with trendy climate change and glib comments (from people who probably know better) about how this all means gloom and doom or even worse.

Evolution of a cyclone. Weak low pressure along a front (a) is enhanced by upper divergence as a jet stream trough (dashed lines) sweeps through the region (b). The cold side of the storm eventually catches up with the warm side (d), and the cyclone dissipates. Shaded regions denote precipitation. This entire process usually takes about three days.

CYCLONE COSTS AND IMPACTS

Whether they know it or not, societies have adapted to rhythm of cyclones beat out of the jet stream. As the jet, and its attendant cyclones migrated from the southern Great Plains to Canada each summer, so did many of the nomadics of pre-European North America. South of the cyclones initiated by the jet, rainfall becomes sporadic and undependable, resulting in an equally spotty distribution of green grass and four-legged chow. North of there, the buffalo did roam.

One would think that features as striking and distinct as cyclones would have provoked some academic to spend a life calculating their costs and benefits to society. But, nooo! Instead, we have only a smattering of disconnected studies, usually financed or directed towards special interests such as the insurance industry or agriculture.

Agriculture

Fall-planted grains, particularly winter wheat, is a cyclone-dependent crop. In Virginia, the average planting date is in October in the west and early November in the east--long after the major rain-producing mechanism switches from summer type thunderstorms and tropical systems to garden variety cyclones. The same is true for the large region stretching from Nebraska and North Texas to the Colorado Front Range. There, the crop is planted a bit earlier (September and October), but the dependence upon cyclone-produced precipitation is the same.

People in this Office have done and published a bit of academic (i.e.-hard to read) research on this subject. We found that in the Great Plains almost two-thirds of the component of winter wheat yield that is due to precipitation is a result of rain and snow produced primarily by cyclones. We suspect that a similar result applies in Virginia.

In contrast, the other two major U.S. crops, corn and soybeans, are not nearly as dependent upon cyclone-produced rains, because they are planted at or near the end of the spring transition from cyclones to summer thunderstorms. Probably less than 20% of the amount of crop yield that can be statistically related to rainfall is generated by cyclones.

Insurance

The insurance industry is interested in cyclones because it has been found that the total number of cyclones observed in a given year is strongly related to the amount of damage claims that are paid. We touched on this in our last Advisory, when we highlighted research by David and Stanley Changnon showing that there appears to be a slight decline in the number of cyclones, and the damages per cyclone show no real net change since they were first measured in the 1950s.

On the other hand, recent years have produced some whopping insurance claims from individual storms. The Changnons found that 1990-94 period was especially costly, in large part because of the so-(mis)-called "Storm of the Century" in March 1993, and some pretty spectacular flooding in the Upper Mississippi. The latter was more associated with thunderstorms than with cyclones, but has contributed to the general perception that the weather has been unusually nasty in the last few years.

Some Big Storms

Different cyclones do different things. Most are beneficial. In fact, it would be a very strange world indeed if most living things were not adapted to the most common form of large scale atmospheric disturbance! However, like grade-schoolers, there are a few bad actors that give the whole crowd a lousy name. Here's our list of the three cyclones of the last half of this century that should have been sent to detention:

The Ash Wednesday Storm (March 6-10, 1962)

This March 1962 cyclone, known as the "Ash Wednesday Storm" for reasons that we hope aren't completely opaque, is a landmark storm in modern history because it was the first storm that demonstrated how the rapid buildup of the Atlantic Coast was creating enormous property exposures. In 1962 dollars, it caused $200,000,000 worth of damage over a broad swath from North Carolina to Cape Cod.

No one has calculated what the cost of this storm would be today, but it's surely not just $200,000,000 times the compounded inflation rate over the last 30 years (even though that figure is a tidy sum!). What we do know is that property values are WAY up. Most of the beach chacks that got washed away in '62 turned into beachfront palaces, if not condos, thirty years later. Roger Pielke Jr., recently estimated that the amount of insured property along the Atlantic Coast is now $3,100,000,000,000 (a really tidy sum!!), or nearly half of the U.S. annual Gross National Product.

This particular storm caused much of its damage because of its long duration and the fact that it occurred during "perigean spring tide," which is the highest average tide in the solar-lunar cycle. Further, a band of high pressure north of the storm stretched all the way to England, making the east-to-west fetch of wind excessively long. Translation: blow wind over water in one direction for 3,000 miles and then have it all come ashore during the highest tide, and you have one very destructive Nor'easter.

The Ash Wednesday storm also holds many of the 20th Century single-storm and 24-hour snowfall records in piedmont Virginia, including Charlottesville and Big Meadows in Shenandoah Park. It did lose a few, such as Lynchburg and Roanoke to the (also-misnamed) "Blizzard of 96."

The Tornado "Super Outbreak" of 1974

Staunton, Virginia has the unlikely distinction of being the site of the last of a remarkable 148 tornadoes spawned by a strong cyclone in the upper midwest on April 3-4, 1974. This remains the largest single outbreak of tornadoes recorded in the world, although it is certainly true that similar events occurred before there were enough people and weather types around to notice.

What made the 1974 outbreak so stellar was that the parent cyclone was unusually strong and situated along a front that was parallel to almost the entire gulf coast. When the storm spun up and the warm front advanced northward, a tremendous amount of gulf moisture surged forward at an extremely rapid rate. This made the atmosphere explosively unstable, forming three simultaneous lines of severe thunderstorms. As the thunderstorms grew in height, they were increasingly spun by the parent cyclone and the jet stream, resulting in tornadoes.

The moisture movement was truly phenomenal. At at 7AM EST, the boundary of the Gulf of Mexico moisture was located from Southern Illinois to the North Carolina-South Carolina border. A mere 8 hours later, it ranged from southern Wisconsin to Philadelphia. This means that the Gulf boundary was moving forward at over 40mph, thanks to the strength of the parent cyclone. And, thanks to this moisture, the last tornadoes in the outbreak were in Virginia.

The "Storm of the Century" (March 12-14, 1993)

We think the title's a bit overblown, but this March, 1993 storm was a whopper. It was sufficiently bad that, even though it was at the end of a remarkably wimpy winter, most people wound up thinking that the entire season was just awful.

It was a record setter for low barometric pressure at several locations in the Mid-Atlantic and western New England. Even Norfolk, which most people would assume ties it's record low pressure to a hurricane, came in with an all-time figure (28.54 inches). Richmond set its record at 28.55, and record lows were also hit at Lynchburg and Dulles Airport. In other words, just about every long-standing record in Virginia fell. But low barometers don't necessarily cause blizzards or newsworthy destruction, unless the associated winds occur in the right place, i.e.-where there are plenty of television cameras handy .

With regard to "mediagenicity," the 1993 cyclone hit the jackpot in spades. It started off with a surprise pounding of the Cuban sugar crop, then brought hurricane force winds and hurricane-like storm surges over Florida. Also, the storm brought a very UN-hurricane like six inches of snow to North Florida, which is the greatest amount in the 20th century (there were bigger totals in 1899 and 1895).

What happened? Some scientists think that an anomalously warm Gulf of Mexico, tepid in part because of the aforementioned wimpy winter, fired up a bunch of thunderstorms around the nascent low pressure system, which, because of their upward motion (thunderheads build "up"), increased the convergence of air at the surface. This blew a cyclone up very far south and west into the Gulf, which itself provided more juice, it being a very warm body of water, which provided more upward motion, which provided an infinity of television videotape.

At least 243 deaths were attributed to the storm, a mortality over three times the total from Category 4 hurricanes Andrew (1992) and Hugo (1989), even though these two storms struck near large centers of coastal population. David Changnon, from Northern Illinois University, estimates the damages from this single storm were $2,800,000,000, making it the most expensive (non-hurricane) cyclone in the history of the nation. But because of the explosion in coastal development since 1962, it's pretty hard to say that the 1962 Ash Wednesday storm would have been much different if it had hit in the early 1990s.

The problem is that the original cold air often is draining away from big high pressure systems located over New England or southeastern Canada. Air diverges away from this cold pool but can't get much further than the mountains, which "dam" the cold air to a depth of about 4,000 feet, which is the average of the highest mountain ridges that run the entire western portion of the state.

In addition, the temperature contrast across the Gulf Stream makes it a kind of "natural" (persistent) front that gets thrown inland as a cyclone develops the the south and west. It then runs up against the cold bubble of air trapped against the mountains. This "pinching" effect creates two, often overlapping zones where the warm air is riding over the cold stuff and enhances precipitation, particularly in the winter. In fact, it is the peculiar correlation between the Gulf Stream, the western mountains, and the cold air streaming down between the two that makes piedmont Virginia an unusually snowy place for its relatively low latitude and altitude.

As shown in another sidebar, the Gulf Stream and the Blue Ridge mountains conspire to put some pretty weird kinks on the normally smooth warm fronts commonly generated by low pressure systems. As a result, warm air from the south rides over cold surface air that is trapped against the mountains, and warm air from the Gulf Stream comes in from the east.

The average depth of the trapped cold air is around 4,000 feet--or as high as the average highest ridgetops in the western portion of the state. This is obviously no accident; but it is a peculiarity of atmospheric physics that this is precisely the depth of cold air that is necessary to create snow--not much less and the best that you can come with is either sleet or sleeze (SLeet and frEEZing rain).

That makes piedmont Virginia a very unusual place for low elevation heavy snow. We took a look around the planet to see how many other places of similar latitude and altitude (37degrees and 500ft, respectively) would see 18 inches or so of snow every decade or so. Outside of the U.S, we couldn't find any. The best candidate, Ashkhabad, Turkmenestan had the altitude but didn't get enough moisture. The remaining possibility, Weifang, China, doesn't have much data available, but the northeast corner of China, including Beijing and Tsingtao, are notoriously dry in the winter.

Almost 20 years ago, State Climatologist Bruce Hayden and his colleague Robert Dolan, an expert on beaches and beach erosion, landed a whale of a contract from the Office of Naval Research to study coastal storms. One of the many things they did was to hire an army of students to pore over old weather maps and to catalog where and when every cyclone-including the daily, nondescript ones-appeared since 1885.

This unique record displayed some interesting changes over timeŅin particular, some long-term oscillations between eras dominated by coastal storms and others notable for their absence.

That was before climate change became trendy, and most of the findings were consigned to the academic oblivion of the stacks of the nation's science libraries. Obviously, though, the intellectual climate has changed, and now everyone, including big insurance companies, wants to know if what humans do to the atmosphere is exposing them to greater coastal risk.

As a result of all this concern, we recently updated the record of cyclone tracks through the mid 1990s, and examined their behavior, particularly in light of dreaded global warming.

A la Hayden, we applied some mathematical wizardry that breaks down the cyclone record into its most likely patterns. It does so in descending order, from the one that explains most of the history of the 110-year record, to the next most important, etc....

Reading these maps is a bit tricky. Think about the numbers as indicating "relative" frequency within a given map. The most important pattern, shown in our picture, displays both positive numbers (off the Atlantic Coast) and negatives (over the midwest and Ohio Valleys). What this means is that the most likely pattern is one in which EITHER there are a large number of Atlantic storms and few in the midwest OR the opposite-a large number of midwestern storms and few in the Atlantic.

Most climate buffs, and/or people who were lucky enough to comprehend our incomprehensible last article, will recognize that this must mean that there are some preferential locations for the jet stream troughs that initiate cyclones-either they're out in western North America or here in the east.

That's actually been known for some time, either anecdotally ("how come it always seems that when the East Coast is cold, the West Coast is hot?") or analytically, where it has been known for decades as the "PNA pattern", for "Pacific-North American" oscillation.

Now for the "cool" part: Every year, the numbers and positions of cyclones determine the strength, or index of this pattern. In years when the index is high, regions on the map that have positive numbers are likely to have above normal numbers of cyclones and vice-versa for the negative zones.

We can plot these values out on a year-to-year basis. For the most important pattern, there is a propensity of negative values in the second half of this century. As noted in our tortured prose, this means that, in general, there have been relatively more Atlantic storms and fewer in the Ohio Valley.

The next most likely pattern is for a winter to simply display a large or small number of coastal storms, with little change elsewhere in eastern North America. The time-history of this one shows nothing special, except a sharp uptick at the end.

SIGNS OF DREADED GLOBAL WARMING?

As noted in our earlier story, the temperature contrast between polar cold and tropical warmth is what fuels the jet stream, and it is the jet that is most instrumental in fanning up cyclones. When the temperature contrast is low, the jet is lazy, headed for its usual summer vacation in Canada, and storms are weak.

Everyone concedes that a greenhouse effect warming of the planet must primarily affect the coldest airmasses at the expense of the warm ones (see box). So, as a general rule, a greenhouse warming should be accompanied by a decline in either storm strength or frequency, or some combination thereof.

Can we find any relation between the patterns of cyclones in eastern North America and temperature change in the 20th century?

The second principle component of winter cyclone tracks.

(As the audience expects yet another negative Advisory result....) Yes we can! We do find a statistically significant positive association between trends in the most important pattern of cyclone distribution (the Atlantic-Midcontinent oscillation) and hemispheric mean temperature.

Now the hard part: explaining. We're going down to the National Climatic Data Center to present this finding at a meeting where the insurance industry is seeking information on "extreme events" (read: can you give us an reason, called climate change, to raise our rates?). While space keeps us from providing Advisory fans with the tedious (and tortured) analysis, it looks like human-induced warming has little if anything to do with this because the much of the warming pattern associated with the change in cyclone tracks took place before humans could have caused it. And during the period of major greenhouse gas emissions-in the last 40 years or so-there's been no real change.

In other words, most of the change in the cyclone patterns evolved in the middle of this century, from 1930 through the mid 1950s. Too bad its not our fault.

Relationship between Northern Hemisphere temperature departures and the first principle component of winter cyclone tracks.

*********

SATELLITE UPDATE

It's been about 15 months since Advisory fans have been bludgeoned with global temperature data from the Microwave Sounding Unit (MSU) satellite orbiters. Since then there have been two big stories:

Global Cooling Now Statistically Significant

That is not a misprint. When averaged over the planet since the record begins on January 1, 1979, the MSU now shows a statistically significant cooling trend. This means that there's only a one-in-20 chance that the downward trend results from the behavior of random (trendless) numbers. Note that graphics in the Virginia Climate Advisory only have trend lines drawn through them when the they are statistically meaningful, unlike a few other (Ahem!) government publications that we know.

This is getting a bit embarrassing. The satellites are now in their 19th consecutive year of measurement. Climate models that served as the basis for the United Nations treaty to ameliorate climate change say there should have been about 0.6C of warming during this period. Newer models that incorporate the cooling from other industrial emissions as a counteraction to the warming from the greenhouse effect, still predict significant warming of around 0.4C.

What does the United Nations' "Consensus of Scientists" choose to do about this in it's new "Policymakers' Summary" on global warming? Ignore it! So, if you look at the entire chapter of their new climate treatise (and the only section anyone who is not a climate nerd will bother to read), you won't find one mention of the word "satellite."

"Silver Bullet" duds out

Because of the intense politicization of the global warming issue, it shouldn't surprise anyone that millions of public dollars are being spent trying to find something "wrong" with the satellite data. Even as they are ignored in the U.N.'s "Policymakers' Summary," they form a very powerful argument against the forecasts of dramatic warming.

Global warming moderates have been remonstrated for several months that the "silver bullet" was about to shatter the satellite data, in the form of an article slated to appear in a spring issue of Nature. We were shaking.

Writing in that prestigious journal, federal climatologists J.W. Hurrell and K.E. Trenberth plugged observed surface temperatures into a computer climate model, and when the model said that the satellite data should be warmer, argued that THEREFORE the satellite data are wrong.

Silly us. And we thought that the purpose of data was to check whether or not computer models are correct. Maybe only in global warming is the normal logic of science suspended!

At any rate, contrary to popular perception (and hence, here in The Current Wisdom) Hurrell and Trenberth did not claim that there was a significant warming in the satellite data. The closest they came was in the last paragraph of the paper, which alleged that, based upon their work, they thought it likely that the lower atmospheric satellite record should more resemble that for the next layer up, which shows a warming of 0.02C per decade. That's right: 1/50 of a degree per ten years. Needless to say, that's statistically insignificant. Some silver bullet.

Reference: Hurrell, J.W., and K.E. Trenberth, 1997. Spurious Trends in Satellite MSU Temperatures form Merging Different Satellite Records. Nature 386, 164-167.

The spatial distribution of trends in the satellite-based temperature record.

---

|Contents | VCA Homepage| Va State Climatology Office Homepage|

---

NWS 90-Day Outlook

• July-September
• August-October

NWS Forecast Verification

• January-March
• February-April
  
  
  

  

  VCA home page
  

  State Climatology Office home page