01 Jan 2008 Hurricane Forecasters’ 0 for 2 Record Reveals Limits of Climate Science, by David Ridenour
With the 2007 hurricane season officially over, we now have a better idea of what 85%-95% forecast certainty means: Wrong 100% of the time. That, at least, has been the case over the past two years.
In August, the National Oceanic and Atmospheric Administration (NOAA) forecast an 85% probability there would be an “above normal” hurricane season.1
This was the second year running the government hurricane forecast was wrong. This 0-2 record may tell us something about other similarly “certain” forecasts, such as those issued by the U.N.’s Intergovernmental Panel on Climate Change.
If forecasters can’t get hurricane projections right in the middle of a hurricane season, updating data constantly, how can we trust their forecasts for a hundred years from now?
NOAA’s forecast called for seven to nine hurricanes, three to five major hurricanes, and 13-16 named storms. According to the agency, the “normal” number of such storms is six hurricanes, two major hurricanes, and 11 named storms.2
There were six hurricanes during the season, only two of which were classified as “major,” category three or more on the Saffir-Simpson scale.3
The sixth hurricane came only three days before the official end of the hurricane season when NOAA’s National Hurricane Center (NHC) quietly upgraded tropical storm Karen to hurricane status.4 The timing of the re-designation – at the moment hurricane season post-mortems were already running in newspapers throughout the country – may have struck some as a bit suspicious.
But National Hurricane Center spokesman Dennis Feltgen insists that the timing was purely coincidental.
“Its re-classification was not related to any particular time of the season,” said Feltgen. “NHC specialists are consummate professionals and would never name (or not name) a tropical cyclone for the sole purpose of verifying a product.”5
He was not able to cite another example, however, when such a re-classification came so close to the end of the season.
Even with Karen’s change in status, the hurricane season was unusually quiet. The Accumulated Cyclone Energy (ACE) Index, which measures storm duration and intensity, was just 70.54 this year,6 more than 25% below the average7 and a little over half of what NOAA predicted. Hurricanes Lorenzo and Humberto both had lower ACEs than any tropical storm in 2007, and the season’s overall ACE – including the two category 5 hurricanes – was still less than the ACE of 2004’s Hurricane Ivan all by itself.8
The only number that NOAA had right was “named storms,” which totaled 15 this year.9 Unfortunately, “normal” doesn’t mean a great deal.
Forecasting “named storms” is a bit like forecasting results of Rorschach tests — those famous ink blot tests. “Named storms” represents what analysts see, not necessarily what is actually there.
Only storms identified as subtropical storms, tropical storms or hurricanes at the time they are occurring are named. At the end of last year, for example, NOAA identified a previously-undesignated tropical storm that had occurred in mid-July after new data, or a new interpretation of data, changed its analysis. That storm was never named.10
The difference between named storms and actual storms can, at times, be significant. In 1964, for example, three tropical storms weren’t named – equal to one-third of named storms that year.11
The criteria for naming storms has also changed – most recently, in 2002, when subtropical storms started to be named along with tropical storms and hurricanes. Not even hurricanes were officially named before 1950.12
Since subjectivity is built into “named storms” it’s a very peculiar thing for NOAA to be forecasting.
Last year, NOAA projected that there would be 12-15 named storms, 7-9 hurricanes and 3-4 major hurricanes and indicated that there was a 95% probability that the hurricane season would be either normal or above normal.13 The agency’s forecast wasn’t even close. There were just nine named storms, five hurricanes and two major hurricanes.14
All this makes one wonder about the level of certainty bandied about by the U.N. Intergovernmental Panel on Climate Change (IPCC).
The IPCC says it is very likely (greater than 90% certain) that the planet will experience greater frequency of warm spells and heat waves over most land areas in the 21st Century, increased precipitation in high latitudes, and slowing of the meridional overturning circulation, the process whereby the ocean’s warm upper waters are transported to the far north and cool deep waters are returned toward the equator resulting in more moderate climates.15
If the IPCC’s projections are anything like NOAA’s, we truly have nothing to worry about.
Global warming alarmists have exploited NOAA’s and the IPCC’s forecasts to frighten the American people in hopes of increasing public support for economically-ruinous caps on U.S. carbon emissions.
And they’re getting plenty of help from NOAA these days not only through these annual forecasts but through reports of new “records” – in named storms, temperatures, and the like – being set that the agency knows it has no way of knowing.
The count of tropical storms (named or otherwise), for example, would have increased dramatically whether actual storm activity had increased or not.
Our ability to monitor weather and tropical cyclones is constantly improving. In the past decade alone, we’ve launched numerous new satellites such as GEOS-10, GEOS-12,16 Quick Scatterometer,17 NOAA-15 through NOAA-18 (all launched within the past decade).18
The Quick Scatterometer, a satellite that measures the ocean’s surface winds launched in 1999, produced more than double the number of daily measurements of its predecessor, the NASA Scatterometer (NSCAT). NSCAT, introduced in 1996, had increased by more than 100 times the amount of ocean wind information previously available from ships.19
Increased resources and technological advancement across the board – from that used in satellites to Hurricane Hunter aircraft – means we are identifying storm systems that we simply wouldn’t have known about in the past. We’re certainly identifying many more than we could when our principle forecast instrument was a wet finger held aloft.
Hurricane Karen is a good case in point. NHC estimates that Karen reached a maximum speed of 65 knots per hour ( or 74.9 miles per hour) – only about 1.2 mph above tropical storm status. No direct measurement of the storm’s wind speed was taken. Instead, an estimate was made based on satellite imagery and data from Hurricane Hunter aircraft using state-of-the-art Stepped-Frequency Microwave Radiometer (SFMR) technology taken six hours after Karen started losing strength.20 The Hurricane Hunter aircraft had only been equipped with the SFMR this summer.21
Even if NHC’s extrapolation of the data is correct – and 1.2 mph leaves little room for error – it’s unlikely that Karen would have been classified as a hurricane in year’s past.
NHC’s Feltgen ducked a specific question on this, but indirectly conceded that it was possible.
“Technology such as satellites and the SFMR has allowed forecasters to detect … several tropical cyclones that might have otherwise gone undetected in the past,” Feltgen said. “In addition, the technology also permits a more accurate measurement of the storm intensity.”22
NOAA makes no mention of its changing criteria for naming storms nor the vast differences in data quality when it issues the annual hurricane season forecasts. Yet it continues to cast these forecasts and subsequent storm reports in a historical context, as though this holds some meaning.
It holds no meaning in science, but in politics.
If NOAA continues to dabble in politics, Americans should afford it all the respect it does to politicians…
…None at all.
David A. Ridenour is Vice President of The National Center for Public Policy Research.
1 “NOAA Updates Atlantic Hurricane Season Outlook: Above-Normal Season Still Expected,” National Oceanic and Atmospheric Administration, August 9, 2007 (available November 21, 2007 at http://www.noaanews.noaa.gov/stories2007/s2905.htm)
3 “Tropical Weather Summary,” November 30, 2007 (available at http://www.nhc.noaa.gov/archive/2007/tws/MIATWSAT_nov.shtml).
4 Richard J. Pasch, “Tropical Cyclone Report: Hurricane Karen,” National Hurricane Center, November 27, 2007 (link available on November 30, 2007 at http://www.nhc.noaa.gov).
5 Email response from Dennis Feltgen, National Hurricane Center, December 7, 2007.
6 “2007 Atlantic Ocean Tropical Cyclones,” National Climatic Data Center (available on December 7 at http://lwf.ncdc.noaa.gov/oa/climate/research/2007/2007-atlantic-trop-cyclones.html).
7 Christopher Landsea, “Tropical Cyclone Records: E11,” Atlantic Oceanographic and Meterological Laboratory’s Hurricane Research Division, June 1, 2007 (available on December 7, 2007 at http://www.aoml.noaa.gov/hrd/tcfaq/tcfaqE.html).
8 “2004 Atlantic Ocean Tropical Cyclones,” National Climatic Data Center (available on December 7 at http://lwf.ncdc.noaa.gov/oa/climate/research/2004/2004-atlantic-trop-cyclones.html).
9 “2007 Tropical Cyclone Advisory Archive,” National Hurricane Center (available on December 11, 2007 at http://www.nhc.noaa.gov/archive/2007/index.shtml).
10 Eric S. Blake and John L. Beven, “Unnamed Tropical Storm (AL022006): 17-18 July 2006,” National Hurricane Center, December 15, 2006 (available November 21, 2007 through link at http://www.nhc.noaa.gov/2006atlan.shtml)
11 Eric S. Blake, Edward N. Rappaport, and Christopher W. Landsea, “The Deadliest, Costliest, and Most Intense United States Tropical Cyclones from 1851 to 2006,” National Weather Service, National Hurricane Center, April 15, 2007 (available on November 21, 2007 at http://www.nhc.noaa.gov/Deadliest_Costliest.shtml).
12 Mark DeMaria, et. al, “Further Improvements to the Statistical Hurricane Intensity Prediction Scheme,” American Meteorological Society, August 2005 (available on November 21, 2007 at http://rammb.cira.colostate.edu/resources/docs/Further_DeMaria.pdf)
13 “NOAA: August 2006 Update to Atlantic Hurricane Season Outlook,” National Oceanic and Atmospheric Administration, August 8, 2006 (available on November 21, 2007 at http://www.cpc.ncep.noaa.gov/products/outlooks/hurricane2006/August/hurricane.shtml).
14 “2006 Atlantic Hurricane Season,” National Hurricane Center (available on December 7, 2007 at http://www.nhc.noaa.gov/2006atlan.shtml).
15 “A Report of Working Group 1 of the Intergovernmental Panel on Climate Change: Summary for Policymakers,” February 2, 2007 (available on November 21, 2007 at http://ipcc-wg1.ucar.edu/wg1/Report/AR4WG1_Print_SPM.pdf).
16 “GOES Next: Geostationary Operational Environmental Satellites Next Generation,” Florida State University (available on November 21, 2007 at http://www.met.fsu.edu/explores/satellites/Goes-next/index.html).
17 “Quick Scatterometer (QuikScat),” NASA (available on November 21, 2007 at http://science.hq.nasa.gov/missions/satellite_51.htm).
18 “The Advanced Next-Generation Television and InfraRed Operational Satellites,” Florida State University (available on November 21, 2007 at http://www.met.fsu.edu/explores/satellites/Atn/index.html).
19 Douglas Isbell, Mary Hardin, Hideo Hasegawa and Hiroyuki Ikenono, “First Wind Data From Scatterometer Captures Pacific Typhoons,” October 3, 1996 (available on November 21, 2007 at http://www.eorc.jaxa.jp/hatoyama/etc/images/adeos_first/nscat_first_e.html).
20 Richard J. Pasch, “Tropical Cyclone Report: Hurricane Karen,” National Hurricane Center, November 27, 2007 (link available on November 30, 2007 at http://www.nhc.noaa.gov).
21 “USAF Upgrading Hurricane Hunter Plane,” UPI, May 11, 2007 (available November 22, 2007 at http://www.upi.com/Security_Terrorism/Briefing/2007/05/11/usaf_upgrading_hurricane_hunter_plane/7148/).
22 Email response from Dennis Feltgen, National Hurricane Center, December 7, 2007.