Category 6™

Postcards III: Robot aircraft for hurricane research

By: JeffMasters, 3:46 PM GMT on April 30, 2008

I'm in Orlando this week for the 28th Conference on Hurricanes and Tropical Meteorology, sponsored by the American Meteorological Society. The conference, held once every two years, brings together the world's experts on hurricane science.

Robot aircraft for hurricane research
The age of Uncrewed Aerial Vehicles (UAVs) for hurricane research is upon us. NASA promoted the use of their Global Hawk UAV in a talk Monday here in Orlando, and yesterday, talks about another UAV--the aerosonde--were delivered by Joe Cione and Guy Cascella of NOAA. Unlike the Global Hawk, the aerosonde has already made flights into hurricanes--into Category 1 Hurricane Ophelia in 2005, and into Category 1 conditions in the extratropical version of Hurricane Noel in 2007. The aerosonde can monitor conditions near the sea surface, where it is too dangerous to fly crewed aircraft. Data taken near the sea surface are particularly important for determining if a hurricane may undergo rapid intensification. The researchers propose to fly the Aerosonde as low as 200 feet inside a hurricane.

Figure 1. The Aerosonde in flight. Image credit:NASA.

The aerosonde is a $50,000 propeller-driven airplane with a 1.6 horsepower engine, made by Aerosonde, based in Melbourne, Australia. The plane can fly 2300 miles on 1.5 gallons of fuel at a cruising speed of 60 mph--good enough for a 20 hour flight. In 2007, an aerosonde spent 17.5 hours in 'Noreaster Noel off the U.S. East Coast, after the hurricane had transitioned to a powerful extratropical storm.

The aerosonde has not been given clearance by the Federal Aviation Administration (FAA) to fly near the U.S. coast, because of concerns that the robot aircraft might encounter other aircraft skirting the storm, according to a Florida Sun Sentinel article posted earlier this year. The FAA does plan to give the aerosonde clearance late in 2008--with restrictions. So, for this year, the aerosonde will fly out of Barbados, where it has been given approval to fly. A 2-week research project is scheduled beginning September 1. So far, the aerosonde has not flown into anything stronger than a Category 1 storm, and the researchers are eager to test the aircraft in a "real" storm this year. The aerosonde will also be "on call" for rapid deployment anywhere in the Caribbean during the period September 15-October 31.

Jeff Masters

Postcards II: hurricane database issues, and the Bill Gray show

By: JeffMasters, 5:24 PM GMT on April 29, 2008

I'm in Orlando this week for the 28th Conference on Hurricanes and Tropical Meteorology, sponsored by the American Meteorological Society. The conference, held once every two years, brings together the world's experts on hurricane science. A few snapshots from the past 24 hours:

Northern Hemisphere tropical cyclone inactivity in 2007
Ryan Maue of Florida State University showed that tropical cyclone activity in 2007 the Northern Hemisphere (Atlantic, Eastern Pacific, Western Pacific, and North Indian Oceans) was at its lowest level since 1977. The Accumulated Cyclone Energy (ACE) is calculated by summing the squares of the estimated maximum sustained velocity of every active tropical storm (wind speed 35 knots or higher), at six-hour intervals. The numbers are usually divided by 10,000 to make them more manageable.

For the Atlantic in 2007, the season's ACE index was 68, 31% below the average of 96. For comparison, the Hurricane Season of 2005 had an ACE of 248. A single storm of the 2004 hurricane season--Hurricane Ivan--had an ACE of 70, more than the ACE index for the entire 2007 hurricane season. For the Northern Hemisphere as a whole, 2007 had the lowest ACE value since 1977. The North Indian Ocean was the only Ocean basin in the Northern Hemisphere that had an above-average ACE in 2007:

Ocean Basin....2007 ACE..Average ACE..%change
Atlantic .............68.........96.........-31%
E. Pacific............52........132.........-60%
W. Pacific...........209........303.........-31%
N. Indian.............44.........16........+275%
N. Hemisphere..376........551.........-31%

Uncertainties in the hurricane data base
HURDAT, the official Atlantic hurricane database, has many significant errors that are slowly being corrected, thanks to a major re-analysis effort being led by Dr. Chris Landsea of the National Hurricane Center. For example, HURDAT lists the Great New England Hurricane of 1938 as a Category 2 extratropical storm (85 knots) at landfall on Long Island, when it was really a Category 3 hurricane with 105 knot winds. Dr. Landsea presented the results of the re-analysis of 12 major hurricanes that hit the U.S. Nine of these 12 storms, including the 1938 hurricane, were re-analyzed to have higher winds at landfall. Dr. Landsea cautioned that our knowledge of past storms in the 1950s and 1960s is quite poor, compared to current capabilities. This occurs despite the fact the hurricane hunters were flying. For example, Hurricane Wilma of 2005 had 280 measurements of its maximum intensity, while Hurricane Carol of 1954, a Category 3 storm that hit North Carolina, had only seven. There were hurricane hunter flights into Carol, but they usually did not fly into the eyewall. It was common for the hurricane hunters to only get close enough to the center to estimate the position using radar back in those days. Carol stalled off the coast of North Carolina for four days, during which time no hurricane hunter flights penetrated into the eye. Could Carol have intensified into a Category 5 storm during that time? We'll never know. Because of such uncertainties, making estimations of trends in Atlantic hurricanes based on HURDAT is difficult to do, Dr. Landsea cautioned.

The Bill Gray show
Dr. Bill Gray of Colorado State University, as usual, generated the most laughter. He commented that he had been to all 28 of these AMS hurricane conferences, with the exception of the first (he was in grad school) and the fourth, when he was in Tokyo. Dr. Gray presented an educational talk, emphasizing the role of natural decades-long cycles in the salinity changes in the Atlantic as being the primary driver of observed increases in Atlantic hurricane activity in recent years. He showed that during 1945-1969 (25 years), during a period the globe was cooling slightly, there were three times as many intense hurricanes in the Atlantic compared to the 25 year period 1970-1994--a period the globe warmed significantly. His tongue-in-cheek conclusion: "CO2 gets into these storms and squashes them!" Extending this result to landfalling U.S. hurricanes, one could claim that we should expect zero landfalling U.S. hurricanes by 2050. Dr. Gray cautioned that this ridiculous result showed that one can manipulate statistics to show virtually any result you want.

Figure 1. Tounge-in-cheek misuse of statistics by Bill Gray to show that the historical record of U.S. landfalling hurricanes predicts zero landfalling hurricanes in the U.S. by 2050 as a result of increasing CO2 in the atmosphere.

Contribution of increases in SST to Atlantic hurricane activity
Adam Lea of University College London presented results showing that the 0.27°C increase in Sea Surface Temperature(SST) between 1996 and 2005 in the Main Development Region (MDR) for hurricanes in the tropical Atlantic was responsible for a 40% increase in hurricanes and intense hurricanes in the Atlantic.

The benefits of hurricanes: rainfall in the Southeast U.S.
David Knight of the University of Virginia showed that hurricanes and tropical storms form an important part of the water budget in the Southeastern U.S. For example, up to 15% of the total rainfall in eastern South Carolina and North Carolina during the six months of hurricane season (June-November) was due to tropical storms or hurricanes between 1980-2004. These numbers are 10-14% for Florida, and 8-10% for Atlanta.

More postcards tomorrow!

Jeff Masters

Postcards from the Orlando Hurricane Conference

By: JeffMasters, 4:41 PM GMT on April 28, 2008

I'm in Orlando this week for the 28th Conference on Hurricanes and Tropical Meteorology, sponsored by the American Meteorological Society. The conference, held once every two years, brings together the world's experts on hurricane science. A few snapshots from this morning's talks:

HWRF hurricane model improvements for 2008
Naomi Surgi of NOAA oulined the progress with the new HWRF model, which debuted last year. The HWRF model outperformed the GFDL model in the Atlantic last year for forecasting hurricane tracks. The GFDL had consistently been the best-performing model for forecasting hurricane tracks in recent years, so this is good news. The HWRF model is intended to eventually replace the GFDL model. However, neither the HWRF or GFDL model performed as well as the GFS model last year, so there is room for the HWRF to improve. New for 2008 for the HWRF model is the ability to include real-time Doppler radar data from the NOAA P-3 hurricane hunter aircraft for initialization of the model. There are also upgrades to the equations governing the model's physics, plus improvements in how the model is initialized. These improvements should make for much improved intensity forecasts beyond 48 hours, Dr. Surgi showed. That's good news, because intensity forecasting has shown very little improvement over the past 15 years, despite a near doubling in the improvement in track forecasts.

Dust from Africa
Amato Evan of the University of Wisconsin showed that dust blowing off the coast of Africa has a very strong impact on Sea Surface Temperatures (SSTs) over the tropical Atlantic, amounting to 0.1°C to 1.0°C annually. Dr. Evan showed that most of the warming of tropical Atlantic SSTs in recent years can be explained by variations in the amount of dust coming off the coast of Africa. In particular, 2005 had very little dust, resulting in unusually high SSTs that help lead to the record breaking Hurricane Season of 2005, with its record 28 named storms.

Influence of Gulf Stream Loop Current on Katrina's intensity
Hurricane Katrina explosively deepened when it passed over an unusually far northern extension of the warm Gulf of Mexico Loop Current, and a Warm Core Eddy that had broken off from the Loop Current. Richard Yablonsky and Isaac Ginis of the University of Rhode Island showed that if the Gulf of Mexico Loop Current had been in its normal position, and no Warm Core Eddy had been present, Katrina would have had a pressure more than 20 mb higher and maximum sustained winds at least 20 mph lower. The study was done using the GFDL model.

I'll have more postcards from the Orlando hurricane conference every day this week. One other highlight from this morning: seeing the 5-year old daughter of one the participants entertain herself by setting up a little diorama complete with ponies, unicorns, and fairies on the floor outside of the main session this morning. Who needs Disney World to entertain a kid in Orlando!

Jeff Masters

Second warmest March on record; will La Niña be gone by hurricane season?

By: JeffMasters, 6:47 PM GMT on April 23, 2008

March 2008 was the 2nd warmest March for the the globe on record, according to statistics released by the National Climatic Data Center. Over the Northern Hemisphere, and over all of the globe's land areas, March 2008 was the warmest March in the 128-year global record. Only the presence of a moderately strong La Niña event that cooled ocean waters in the central and eastern equatorial Pacific prevented March 2008 from surpassing March 2002 as the warmest March on record. March broke a string of three straight months when the globe did not record a top ten warmest month ever. Between February 2006 and November 2007, the globe set top ten monthly warm temperature records for 22 straight months.

Figure 1. Departure of temperature from average (the anomaly) for March of 2008, the second warmest March on record for the globe. While the U.S. recorded slightly below average temperatures, much of Asia and Europe saw remarkably warm temperatures. Image credit: National Climatic Data Center.

How much cooling did La Niña give to the globe in March?
La Niña is a periodic cooling of the Equatorial waters of the Eastern Pacific that occurs every 3-7 years. The cooling is due to a natural cycle of anomalous winds from the east that act to push surface waters away from the coast of South America, allowing cold water from deep in the ocean to rise to the surface to replace the surface waters blown to the west. These cool waters often cause a noticeable drop in global temperatures. Conversely, when the opposite phenomena occurs--an El Niño event, which brings anomalously warm Sea Surface Temperatures (SSTs) to the Equatorial Eastern Pacific, enough heat is added to the atmosphere that global temperatures warm significantly. According to Trenberth et al. (2002), a typical El Niño event increases global temperatures by about 0.1°C. Exceptional El Niño events, such as occurred in 1997-1998 and 1982-1983, increase global temperatures by up to 0.2°C. El Niño events heat the atmosphere by causing changes in cloudiness and atmospheric circulation, and through direct radiation of heat from the ocean to the atmosphere. There is a lag of 3-6 months between the time an El Niño event occurs and the time the atmosphere heats in response. Similarly, when a La Niña event occurs, heat is drawn out of the atmosphere and the oceans are recharged with heat. Global temperatures cool, again with a lag of 3-6 months. The correlation between global temperature anomalies and El Niño/La Niña temperature anomalies can be plainly seen in Figure 2. Note that the correlation is not perfect--there are some El Niño/La Niña events that do not affect the global temperature much. For example, global temperatures did not cool much during the strong 1988-1989 La Niña event. Therefore, it is a good bet--but not certain one--that had we not had a strong La Niña event this winter, March 2008 would have been the warmest March on record, since it missed the record by only 0.04°C.

Figure 2. Comparison of temperature anomalies in the equatorial Pacific Ocean bounded by a box between 5°S, 5°N, 120°W, and 180°W (the Niño 3.4 region) and global temperatures for 1950-1998. Means for 1950-1976 and 1977-1998 (horizontal lines) are shown separately to highlight a climate shift that occurred in 1976/1977. The reasons for this shift are unknown. Note that when an El Niño event occurs, the globe tends to warm by about 0.1°C, and when a La Niña event occurs, the globe tends to cool. Image credit: American Geophysical Union's Journal of Geophysical Research.

Warmest month ever stats
It is interesting to compare what the phase of El Niño/La Niña was during each of the 12 record warmest months the globe has recorded. If we adjust for the 3-6 month lag between an El Niño/La Niña event and the monthly global temperature records, it turns out that nine of the twelve monthly records were set when an El Niño event occurred during the 3-6 month period prior to the record. Considering that climatologically El Niño conditions are present only about 25% of the time, El Niño has a major impact on when record global warmth will occur. The warmest year on record, 1998, occurred during the strongest El Niño of the past century. The table below compares the 12 monthly global temperature records with the temperature in the Niño 3.4 region (a 3-month average centered four months before the record was set). El Niño events, which occur when the Niño 3.4 index is greater than 0.4°C, are marked with an "E".

Record......Niño 3.4 index
Jan 2007 +0.7 E
Feb 1998 +1.7 E
Mar 2002 -0.1
Apr 1998 +2.5 E
May 1998 +2.3 E
Jun 2005 +0.5 E
Jul 1998 +1.4 E
Aug 1998 +1.1 E
Sep 2005 +0.5 E
Oct 2003 +0.0
Nov 2004 +0.7 E
Dec 2003 +0.4

Figure 3. A La Niña event exists when ocean surface temperatures in the equatorial Pacific Ocean bounded by a box between 5°S, 5°N, 120°W, and 180°W (the Niño 3.4 region) are cooler than 0.4°C below average (based on means from 1971-2000). La Niña events between 0.5°C and 0.9°C are referred to as weak, 1.0°C and 1.4°C are moderate, and 1.5°C or cooler, strong. The winter of 2007-2008 saw strong La Niña conditions, but this has weakened to a moderate event in March. Image credit: NOAA Climate Prediction Center.

An El Niño by hurricane season?
Presence of El Niño conditions usually causes enhanced levels of wind shear over the Atlantic, reducing hurricane activity, so it would be nice to see an El Niño this Fall. The strong La Niña event we had over the past winter has weakened considerably in the past two months, and is now classified as a moderate event, according to the latest El Niño discussion issued by NOAA's Climate Prediction Center. There is some hope that an El Niño will develop by hurricane season. Two of the long-range computer models are now calling for an El Niño to develop by hurricane season (Figure 4), and none of them were calling for El Niño last month. However, there is probably not time for a full-fledged El Niño event to develop, and it is expected that we will have weak La Niña or neutral conditions this hurricane season. Since reliable El Niño records began in 1950, there has never been a switch over to El Niño by hurricane season from a La Niña as strong as the one we have now. Columbia University's International Research Institute is predicting that neutral El Niño conditions are most likely for the coming hurricane season (57% chance), with a 20% chance of an El Niño, and 23% chance of a La Niña. This is pretty much what climatology says--on average, we experience El Niño conditions 25% of the time and La Niña conditions 25% of the time.

Figure 4. Computer model forecasts of El Niño/La Niña made in April. The forecasts that go above the red line at +0.5°C denote El Niño conditions; -0.5°C to +0.5°C denote neutral conditions, and below -0.5°C denote La Niña conditions. Image credit: Columbia University's IRI.

Trenberth, K. E., J. M. Caron, D. P. Stepaniak, and S. Worley, >Evolution of El Niño, Southern Oscillation and global atmospheric surface temperatures", J. Geophys. Res., 107(D8), 4065, doi:10.1029/2000JD000298, 2002.

I'll be in Orlando next week for the American Meteorological Society's bi-annual hurricane conference, and plan to make some quick posts during the week to update everyone on the latest hurricane research. I may make one more post before then.

Jeff Masters

Climate Summaries

Earth Day photos

By: JeffMasters, 7:02 PM GMT on April 21, 2008

Tuesday, April 22, marks Earth Day. As has been my tradition, on Earth Day I share some of my favorite photos wunderground users have uploaded over the past year. The Earth's atmosphere is a fantastically beautiful and complicated creation, and I thank all of you who took the time to share the spectacular things you've seen with the rest of us!

I'll be back Wednesday to talk about the demise of this year's La Niña.

Jeff Masters

Atmospheric Phenomena

2008 hurricane season outlook

By: JeffMasters, 2:56 PM GMT on April 13, 2008

It's too soon to find many clues about what the upcoming hurricane season might be like, since we are facing the famed "April Predictability Barrier". The atmosphere is not predictable enough to make a skillful forecast of seasonal hurricane activity in April (this changes by late May, when skillful predictions of the upcoming hurricane season CAN be made). Thus, we should put little faith in the predictions by Dr. Bill Gray's Colorado State University group and Tropical Storm Risk (TSR) Inc., made in April, calling for 15 named storms, 8 hurricanes, and 4 intense hurricanes. These numbers are exactly what the climatology has been in the 13 years since the current active hurricane period began in 1995, and are pretty unremarkable forecasts.

Figure 1. Sea Surface Temperature (SST) comparison between April 11, 2008 (top) and April 11, 2005 (bottom). SSTs were at record high levels in 2005, and are much cooler in 2008 over the Main Development Region (MDR) for hurricanes. The red wiggly line is the 26°C isotherm, which is the coolest temperature that hurricanes typically form at. Note that it stays warm enough to support hurricanes year-round in the Caribbean. High levels of wind shear prevent development in the winter and spring. Image credit: NOAA/AOML.

Sea Surface Temperature (SST) Outlook
Sea Surface temperatures are at the heart of any attempt to analyze seasonal hurricane activity, since SSTs change slowly and have a major impact on both the number of storms that form and their intensity. For example, the record-breaking Hurricane Season of 2005 had the warmest SSTs ever observed in the tropical Atlantic. Comparing the SSTs in April of 2008 with April of 2005 (Figure 1), we see that SSTs are more than 1 degree C cooler this year over a large portion of the Atlantic Main Development Region (MDR) for hurricanes. The MDR is where 85% of all major hurricane form. Overall, SSTs this April are near average compared to climatology when averaged over the Main Development Region. There is still time for a significant warming of SSTs to occur if we get a large decrease in the speed of the trade winds over the next few months. This would occur if the Bermuda High weakened to much below-average levels, driving slower clockwise winds around its center. TSR is forecasting only a modestly weaker Bermuda High than average, with trade wind speeds about 1 mph below average between now and hurricane season. This would allow SSTs in the MDR to warm to slightly above-average levels. Both Columbia University's IRI and NOAA's CFS model (Figure 2) are predicting average to slightly below average SSTs over the Caribbean and Main Development region during the peak hurricane season months of August, September, and October. The accuracy of these sort of long-range forecasts is a little better than chance, but not a lot. The bottom line is that SSTs should be near average this hurricane season, and should neither enhance nor inhibit hurricane formation. Note, though, that some of the regions outside the MDR, where the other 15% of major hurricanes form, have SSTs as warm or slightly warmer than 2005. For example, Katrina and Rita formed outside the MDR over the Bahamas, and SSTs there this year are comparable to 2005 levels. SSTs are also quite warm in the Gulf of Mexico this spring. Thus, we can expect plenty of fuel for any hurricane that might get loose in the Gulf of Mexico this hurricane season.

Figure 2. Sea Surface Temperature (SST) departure from average (left) and wind shear departure from average (right) from the April 12, 2008 run of NOAA's CFS model. The forecasts are for the 3-month period August-October spanning the most active part of hurricane season. The model predicts near average SSTs over the tropical Atlantic, and below average wind shear. Image credit:NOAA Climate Prediction Center.

El Niño/La Niña Outlook
Obviously, SSTs don't tell the whole story, since the second highest SSTs in the tropical Atlantic since the 1870s occurred in 2006. That was a very normal year with 10 named storms, 5 hurricanes, and 2 intense hurricanes, because an El Niño event occurred. El Niño events usually suppress Atlantic hurricane activity, by bringing increased wind shear and dryer, sinking air over the Atlantic. This is not always the case--recall 2004? El Niño conditions were even stronger that year (as measured by SST departures from average in the Equatorial Pacific), yet that year saw 15 named storms, 9 hurricanes, and 6 intense hurricanes in the Atlantic. Florida got walloped with four hurricanes.

Presence of La Niña conditions usually causes reduced levels of wind shear over the Atlantic, enhancing hurricane activity. The current strong La Niña event has begun weakening noticeably in the past few weeks, giving some hope that La Niña will be gone by hurricane season. However, there is probably not time for a full-fledged El Niño event to replace it by hurricane season, and it is expected that we will have weak La Niña or neutral conditions this hurricane season. Since the active period of hurricane activity that we are in began in 1995, both La Niña and neutral years have seen very high levels of hurricane activity (Figure 3). In fact, neutral years have had even higher hurricane activity than La Niña years (thanks in great measure to the Hurricane Season of 2005).

The NOAA CFS model (Figure 2) is predicting La Niña conditions and lower than average wind shear for the coming hurricane season. None of the computer models are forecasting a switch over to El Niño conditions this year. Keep in mind, though, that the accuracy of these long range models is poor, particularly for forecasts made in March and April. Still, La Niña is well enough established now that it would be a major surprise to see an El Niño arrive by hurricane season.

Figure 3. Observed numbers of named storms, hurricanes, and intense hurricanes (Category 3 and higher) for the 13-year period beginning in 1995. Background image is of Australia's Tropical Cyclone Monica, the most intense storm of 2006.

African dust outlook
African dust is thought to suppress Atlantic hurricane activity, although its role is not well understood. As I explained in a 2006 blog entry, research shows that the presence of drought conditions in the Sahel region of Africa the previous year will increase the amount of dust wafting over the Atlantic during hurricane season. This occurs because drought-damaged soil takes about a year to dry up and turn to dust that can blow away. Last year saw above-average rains during the rainy season (June-September) over the Sahel (Figure 4). This was also the case in 2005 and 2006, so in theory, three straight years of good rains in the Sahel should act to keep African dust levels over the Atlantic below average this hurricane season. The last significant drought years in the Sahel were 2001 and 2002. I made the same forecast last year, but we saw unexpectedly high levels of dust over the eastern Atlantic in July and August, which substantially cooled the ocean waters by blocking sunlight. Dust levels returned to near average levels in September.

Figure 4. Departure of precipitation from average in Africa for August 2007. The region in the red box is the Sahel region of Africa that accounts for most of the year-to-year variability in dust transport over the Atlantic Ocean. Image credit: NOAA Climate Prediction Center.

The outlook for the 2008 Atlantic hurricane season
If the forecasts of near normal sea surface temperatures, below average wind shear, no El Niño, and below average African dust come true, the 2008 Atlantic hurricane season should be well above average in activity. We are also in the midst of an active hurricane period that began in 1995. Since 1995, we've averaged 15 named storms, 8 hurricanes, and 4 intense hurricanes, and this is a reasonable forecast for 2008 (the 100-year climatology is 10 named storms, 6 hurricanes, and 2 intense hurricanes). By May, the atmosphere and ocean begin to give us significant clues about the upcoming hurricane season. Tune into the early June seasonal forecasts issued by NOAA, the Colorado State group, and TSR!

This will be my last blog until Monday April 21, as I'll be on vacation in Arizona.

Jeff Masters

Atlantic hurricane season forecasts for 2008 released

By: JeffMasters, 5:06 PM GMT on April 09, 2008

A well above average Atlantic hurricane season is on tap for 2008, according to the latest seasonal forecast issued today by Dr. Bill Gray and Phil Klotzbach of Colorado State University (CSU). The Gray/Klotzbach team is calling for 15 named storms, 8 hurricanes, and 4 intense hurricanes. An average season has 10-11 named storms, 6 hurricanes, and 2 intense hurricanes. The new forecast is a bump up from their December forecast, which called for 13 named storms, 7 hurricanes, and 3 intense hurricanes. The new forecast calls for an above average chance of a major hurricane hitting the U.S., both along the East Coast (45% chance, 31% chance is average) and the Gulf Coast (44% chance, 30% chance is average). The Caribbean is also forecast to have an above average risk of a major hurricane.

The forecasters cited several reasons for their forecast of an active season:

1) Above-average sea surface temperatures (SSTs) in the eastern subtropical Atlantic and cooler-than-normal SSTs in the South Atlantic have weakened the Bermuda-Azores High. This has resulted in lower surface wind speeds over the tropical Atlantic, and these weak trade winds are expected to persist into hurricane season. Weak trade winds reduce the amount of evaporative cooling of the ocean, resulting in warmer SSTs and lower surface pressures during hurricane season. Hurricanes like to form in an environment with low surface pressures and high SSTs.

2) Hurricane activity in the Atlantic is lowest during El Niño years and highest during La Niña or neutral years. The current strong La Niña event has begun weakening noticeably in the past few weeks. However, there is probably not time for a full-fledged El Niño event to replace it by hurricane season, and it is expected that we will have weak La Niña or neutral conditions this hurricane season. None of the computer models are forecasting a switch over to El Niño conditions this year (Figure 1). Keep in mind, though, that the accuracy of these long range models is poor, particularly for forecasts made in March and April.

3) We are in the midst of an active hurricane period that began in 1995.

How accurate are the April forecasts?
Today's forecast includes the statement, "These real-time operational early April forecasts have not shown forecast skill over climatology" during the 13-year period 1995-2007. In other words, today's forecast has no skill, and should merely be viewed as an interesting experimental research product. I like the fact that they are trying to make useful seasonal hurricane forecasts, but we should wait until their June 3 forecast before putting faith in their 2008 hurricane season forecasts. The CSU team has posted an Excel spreadsheet of their forecast errors (expressed as a mathematical correlation coefficient, where positive means a skilled forecast, and negative means they did worse than climatology). You can see from their numbers that the December and April forecasts have near zero skill, but the early June forecasts have substantial skill. To rectify their poor April forecast skill, the CSU team is trying a new scheme for this year's April forecast. Hopefully the new scheme will show positive skill forecasting upcoming hurricanes seasons, and not just "hindcasting" the past ones. For now, you're best off just paying attention to their early June forecast, which has been quite skillful over the past ten years.

Figure 1. Computer model forecasts of El Niño/La Niña made in March. The forecasts that go above the red line at +0.5°C denote El Niño conditions; -0.5°C to +0.5°C denote neutral conditions, and below -0.5°C denote La Niña conditions. Image credit: Columbia University's IRI.

2008 Atlantic hurricane season forecast from Tropical Storm Risk, Inc.
The British private forecasting firm Tropical Storm Risk, Inc. (TSR), issued a 2008 Atlantic hurricane season forecast this week as well. TSR has almost the same forecast as the CSU team--14.8 named storms, 7.8 hurricanes, and 3.5 intense hurricanes. Unlike the CSU team, these numbers represent a decrease from their December forecast numbers, which were 15.4 named storms, 8.3 hurricanes, and 3.7 intense hurricanes. I like how they put their skill level right next to their forecast numbers: 7% skill at forecasting the number of named storms, 11% skill for hurricanes, and 10% skill for intense hurricanes. That's not much better than flipping a coin, but it is better than the near-zero forecast skill of the Gray/Klotzbach April forecasts. However, TSR doesn't mention the fact that part of their skill may be due to the fact that they issue forecasts of fractional storms (we're not going to get 7.8 hurricanes this year!) If we round these numbers to whole storms, the TSR skill numbers may decrease.

TSR projects that four named storms will hit the U.S., with 1.7 of these being hurricanes. In the Lesser Antilles Islands of the Caribbean, TSR projects 1.5 named storms, 0.7 of these being hurricanes. TSR cites one main factor for their forecast of an active season: slower than normal trade winds July-September over the Caribbean. Trade winds are forecast to be 0.4 meters per second (about 1 mph) slower than average, which would create greater spin for developing storms, and allow the oceans to heat up due to reduced evaporational cooling. TSR forecasts that SSTs will be near average in the tropical Atlantic during hurricane season.

Figure 2. Accuracy of long-range forecasts of Atlantic hurricane season activity performed by Bill Gray and Phil Klotzbach of Colorado State University (colored squares) and TSR (colored lines). The CSU team's April forecast skill is not plotted, but is near zero. The skill is measured by the Mean Square Skill Score (MSSS), which looks at the error and squares it, then compares the percent improvement the forecast has over a climatological forecast of 10 named storms, 6 hurricanes, and 2 intense hurricanes. TS=Tropical Storms, H=Hurricanes, IH=Intense Hurricanes, ACE=Accumulated Cyclone Energy, NTC=Net Tropical Cyclone Activity. Image credit: TSR.

Jeff Masters

Arctic sea ice melting season begins

By: JeffMasters, 3:24 AM GMT on April 06, 2008

The annual Arctic melting season has begun. We've just passed the Spring Equinox, so the North Pole is now in 24-hour daylight. Will the melting of Arctic sea ice this year surpass last year's record? Well, we have a greater areal extent of ice over the Arctic this month compared to April of last year, thanks to some cool Arctic temperatures this winter. In particular, the ice in the Bering Straight between Alaska and Russia extends quite a bit further south than in 2007. This extra ice will likely delay the melting season a bit this year, giving some hope that we won't surpass last year's record melt. However, if weather similar to last summer occurs--unusually clear skies and high pressure over the Arctic--this extra ice will not help much, because it is all thin, first-year ice. It is the thick, 2-9 year old ice that is most able to resist summer melting, and the amount of old multi-year ice is only about half of what it was in 2007. This is apparent from images taken by the QuikSCAT satellite, which carries the SeaWinds scatterometer. This instrument emits a pulse of microwave energy that bounces off the ice and returns to the satellite. Old, multi-year ice is thicker, and reflects a different amount of microwave energy back to the satellite than thin ice, resulting in a whiter image. Thin, first-year ice appears a darker grey. It is apparent from Figure 1 that we have only about half of the old, multi-year ice that we had last year. In fact, thin first-year ice extends past the North Pole, raising the distinct possibility that this year's melt will allow one to sail a ship all the way to the North Pole in September, for the first time since humans began testing Arctic waters with ships in 1497. In addition, a large region of the old ice north of Alaska is highly fractured, making it vulnerable to melting.

Figure 1. QuikSCAT images of the Arctic from April 4, 2007 and April 4, 2008. The boundary of old, multi-year ice is marked in yellow, and 2008 has about half the old ice of 2007. Image credit: NOAA/NESDIS/ORA.

How did we lose so much old ice in the past year?
Part of the lost old ice melted during the record-breaking melt season of 2007, which was fueled both by global warming and a natural (but unusual) sunny summer. Another big chunk of old ice was lost due to natural wind patterns between September 2007 and March 2008. An animation of the sea ice available from Environment Canada's sea ice page (click on "Updated QuikScat animation") reveals that strong winds pushed large amounts of old ice out of the Arctic southward along the east coast of Greenland. So, we can't blame the melting of the Arctic sea ice entirely on global warming--natural weather patterns also played a significant role.

The forecast
It's impossible to guess what the dominant Arctic weather pattern will be this summer, and what level of melting we will get. With the loss of so much old ice over the past year, though, even an average summer has the potential to melt much more ice this summer compared to last summer--all the way to the North Pole. There's also a good chance that we'll see the fabled Northwest Passage open up again, since most of the ice along the Passage is young, first-year ice. We'll just have to wait and see how the summer unfolds.

Jeff Masters

Climate Change

Cyclopsychic research breakthrough proves hurricanes/global warming connection

By: JeffMasters, 12:16 PM GMT on April 01, 2008

A stunning new breakthrough in hurricane research has conclusively settled the matter: global warming is making Atlantic hurricanes and tropical storms more frequent. The new research, accepted for publication later this millennium in The Journal of Irreproducible Results, offers incontrovertible proof that global warming has increased Atlantic named storms by 57-67% over the past century. Using the pioneering new techniques of cyclopsychic storm detection and psychomortorodentiatempestology, the researchers, Professors Peter Webcaster and Judith Flurryfury of the Georgia Institute of Technophobia, and Dr. Greg Hallmonitor of the Colorado Association for Research and Modeling of the Atmosphere (CARMA), showed unequivocally that the lack of satellite measurements and aircraft reconnaissance in the early part of the hurricane record led to only a modest undercount of Atlantic tropical storms. Thus, more than half of the observed increase in named storms in the past century can be attributed to global warming.

"It's well-known that the number of Atlantic named storms has risen from 7-9 per year 100 years ago to 14-15 per year during the present active hurricane period that began in 1995," commented Professor Webcaster in an interview today. "Some MEEAT-loving hurricane researchers (Measure Everything, Everywhere, All the Time) have claimed that this rise was not real, since satellites and reconnaissance aircraft were not around to detect storms early in the hurricane record. We've made efforts in the past to quantify the number of 'missed' historical Atlantic storms using estimates of historical shipping traffic density, and computer regression models that compare recent well-observed storm activity to past storm activity. However, these studies have been pooh-poohed by the MEEAT men, who refuse to believe any science that comes out of a model. So, I began thinking about how we could actually go about observing historical Atlantic storms that were 'missed'. I began thinking the problem in a new light after watching my favorite episode of Star Trek: The Next Generation, "Timescape", where subspace entity mistakenly lays her eggs in the warp core singularity of a Romulan warship, creating a temporal anomaly that forces time to flow backwards. This inspired me to think outside the box, and it occurred to me that paranormal methods might offer a way to see back in time and make actual observations of past storms--and offer a technophobic solution to the problem, as encouraged by the charter of my university, the Georgia Institute of Technophobia."

Figure 1. Cyclopsychic observations of "missing" Atlantic tropical storms during the 20th century. All observations were performed by trained cyclopsychic Madame Cyclotropia. Note the significant drop in "missed" storms beginning in the 1940s, corresponding to the advent of aircraft reconnaissance, and in the 1970s, when satellite coverage of the Atlantic Ocean began.

Webcaster teamed with Hallmonitor and Flurryfury to experiment with a variety of paranormal techniques to make actual observations of past "missing" storms, using Ouija Boards, crystal ball gazing, the Magic Eight Ball, and channeling of restless dead spirits. Initial experiments proved discouraging, though, when the researchers attempted to perform the study themselves. "We were feeling depressed about how the research was going, having just stayed up late one Friday night in Greg's lab in Boulder trying to get the dang Magic Eight Ball to say something other than just REPLY HAZY, TRY AGAIN LATER," related Dr. Webcaster. "So, we decided to give up for the night and down a few shots of grape jello spiked with grain alcohol and delve into Greg's extensive collection of Zippy the Pinhead comic books. After a few jello shots and Zippy comics, we got feeling pretty loose, and, Yow! Decided to trek down to Pearl Street to check out the weekend psychic fair. Well, we got to staggering around the tents of the psychic fair, belting out the sorrowful lyrics of our own version of "Somewhere over the rainbow" we made up:

Somewhere, over the ocean
Back in time
Cyclones formed and decayed
Unseen by humankind

Somehow, we'll find out how many
Before we die
But it doesn't look good
'Cause the Magic Eight Ball lies!

Suddenly, we saw a mysterious shadowy figure beckoning to us from the entrance of a nearby tent, which was emblazoned with the words, Madame Cyclotropia: Psychic Readings for Troubled Atmospheric Scientists. 'I can help you find your missing storms', the seer in the shadows croaked, 'for I know much that is hidden. Come into my lair, and I will reveal the key to unlocking the mysteries of storms long past'. Greg and I looked at each other, shrugged, walked in her tent, and the rest is history."

Once in Madame Cyclotropia's tent, the researchers quickly realized that their limited scientific training could not hope to allow them to conduct rigorous paranormal research. Only a true cyclopsychic with "The Gift" could see back into the dim mists of time to divine the existence of heretofore unknown tropical cyclones. Using her cyclopsychic gift, Madame Cyclotropia correctly divined the past tracks of numerous known storms the scientists challenged her with. However, when asked to divine the existence of "missing" Atlantic storms that had not made it into the official database, she prophesied that she would only be able to do so if the scientists would write her into their latest grant proposal. This grant proposal would surely get funded, she predicted. The scientists eagerly agreed, and headed back to the lab to work on the new proposal.

Webcaster, Hallmonitor, and Flurryfury's proposal, titled, "Using a Trained Cyclopsychic to Divine Past Atlantic Tropical Cyclone Activity", was submitted to the Foundation for Atmospheric Research for Science and Education (FARSE) in early 2007 and accepted later that year. After receiving their grant money, the scientists began spending long nights in Madame Cyclotropia's tent, documenting her revelations from the four primary cyclopsychic techniques: Ouija Board, crystal ball gazing, the Magic Eight Ball, and channeling of restless dead spirits. According to Dr. Hallmonitor, "We were thrilled when the first three techniques we tried all yielded virtually identical results, showing the robustness of our experimental methodology. The three techniques all showed a noticeable drop in the number of "missed" storms in the 1940s, when aircraft reconnaissance became available, and in the 1970s, when satellites coverage began over the Atlantic Ocean. However, when we tried to channel restless dead spirits, we ran into a roadblock. We couldn't find any restless dead spirits with an interest or knowledge of historical Atlantic hurricanes. We happily attributed this to the propensity of dead meteorologists to wind up inside Heaven's Pearly Gates, but were sad that our research would lack this crucial final proof of its validity. We were about to give up when Peter then hit upon the idea of contacting the spirits of groundhogs, who are known for their weather prognosticating ability. Some of these prognosticating rodents might have unfinished business that would keep their restless souls adrift in the ether, available for consultation on weather-related matters. We coined word psychomortorodentiatempestology to describe this exciting new branch of hurricane science, and set off in search of gifted groundhogs spirits with this special skill."

Figure 2. Wee Willy One and Chucky before their departure into the hereafter. Which rodent's spirit would you trust to get accurate weather information from?

Indeed, Madame Cyclotropia was able to contact the spirit of "Wee Willy One", a famed albino groundhog that had once burrowed under the fair gardens of Wiarton, Ontario, and provided weather forecasts each Groundhog's Day up until his death in 2006. Wee Willy One proved to be testy and uncooperative, though, deliberately delivering incorrect storm information. The researchers sought out help from cyclotherapy experts from the Center for Disease Control's Weather Related Illness Division to determine if cyclotherapy might help Wee Willy One overcome his bad attitude. Cyclotherapist Dr. Sandy Chirpchuckle diagnosed Wee Willy One as a cyclopath suffering from rare form of cyclopsychosis. Ordinarily, cyclopsychosis manifests itself only in hurricane scientists and weather enthusiasts during the long, dull months prior to hurricane season. The despondent victims of cyclopsychosis spend long hours in front of flickering computer monitors in dark, gloomy rooms, obsessively poring over maps and statistics of hurricanes long gone by. The victims tend to become highly antisocial but never violent, and can be successfully treated with cycloactive drugs. However, Dr. Chirpchuckle diagnosed Wee Willy One with an extremely rare case of "shadow" cyclopsychosis, brought on by the cyclological trauma being rudely hauled out of his burrow each February 2 so that a bunch of cockamaimie humans could see whether he saw his shadow or not. "Shadow" cyclopsychosis is incurable, both in this world and the hereafter, so Madame Cyclotropia was forced to seek out other groundhog spirits. After months of effort, she finally found the spirit of "Chucky", a friendly groundhog that had once burrowed under the gardens of Nashville, Tennessee. Chucky eagerly provided accurate information on the "missing" Atlantic tropical storms that was precisely in agreement with the data collected from the other cyclopsychic techniques. "We were ecstatic," exclaimed Dr. Hallmonitor. "More jello shots!"

Hurricane experts world-wide are hailing the new findings. "These exciting results conclusively prove that even us blind squirrels can find some nuts," enthused renown hurricane expert, Dr. Kerry Readthemanual of the Massachusetts Institute of Technophobia. Dr. Readthemanual has been a leading proponent of the global warming/Atlantic hurricane link. Even former critics are praising the new findings. Dr. William Graymatter, Professor Über-Emeritus of Colorado State University's Center for Hurricane Observation, Measurement, and Prediction (CHOMP), said in an interview: "I've been in the hurricane business for 113 years, and I know good research when I see it. The findings of Webcaster, Hallmonitor, and Flurryfury are based on solid observational evidence and white magic. There's no black magic involved, such as the use of a computer model, so their results are impregnable."

Dr. Chris Blandsee, Chief Scientist of the Natural Hurricane Center's division of Global Warming Isn't Responsible for the Recent Upswing in Atlantic Hurricane Activity, and Even If It Was, We Wouldn't be Able to Tell, Since the Quality of the Atlantic Hurricane Database is Too Poor to Use for Such Purposes (NHC/GWIRRUAHAEIIWWWATSQAHDTPUSP), has also been critical of past research showing a link between hurricanes and global warming, maintaining that global warming isn't responsible for the recent upswing in Atlantic hurricane activity, and even if it was, we wouldn't be able to tell, since the quality of the Atlantic hurricane database is too poor to use for such purposes. It was his Congressional testimony, along with that of former NHC director Max Minefield, which inspired President Bushwhacker's administration to rename the National Hurricane Center the "Natural Hurricane Center" last year. (This action was also urged by the Government Anagram Accountability Office (GAAO), which found that the letters in "National Hurricane Center" could be rearranged to spell the ominous phrase, "Errant Herculean Inaction"--and also the disturbing, "Teenier Charlatan Unicorn", and the clearly unacceptable, "Inhale Cocaine, Errant Runt!", while the letters in "Natural Hurricane Center" could be rearranged to form phrases much more in harmony with the NHC mission, such as "Natural, Neater, Crunchier.")

Dr. Blandsee grudgingly gave ground in his comments today. "It looks like Webcaster, Hallmonitor, and Flurryfury (and don't try to say her name three times fast) have done some pretty rigorous scientific work," he conceded. "But they've written what is probably the longest and most excruciatingly dull hurricane science paper of all time. All those old storms and their analyzed tracks that they talk about, on and on and on, year by year by year. Ugh! A lot of good trees died to publish that paper. It was even duller than some of my clunkers!"

What's next for the pioneering researchers? "Well, CARMA and the Georgia Institute of Technophobia are collaborating on a grant proposal with Dr. Graymatter and Phil Flossblack of CHOMP to apply cyclopsychic methods in a new way--improvement of seasonal hurricane forecasts," said Dr. Flurryfury. "We've submitted a proposal to FARSE titled, 'Gray Magic: Using Cyclopsychic Methods to Improve Seasonal Hurricane Forecasts'. Lord knows, the forecast busts of the past two hurricane seasons have shown that Flossblack and Dr. Graymatter could use some supernatural help with their predictions."

April Fools!
Meff Jasters

Hallmonitor, G.J., and P.J. Webcaster, 2007, "Heightened tropical cyclone activity in the North Atlantic: natural variability or climate trend?" Philosophical Transactions of The Royal Society A 365, Number 1860, 15 November 2007, Pages: 26952716 DOI: 10.1098/rsta.2007.2083


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

Category 6™


Cat 6 lead authors: WU cofounder Dr. Jeff Masters (right), who flew w/NOAA Hurricane Hunters 1986-1990, & WU meteorologist Bob Henson, @bhensonweather