Tag Archives: gach

Weather Radar in Hawaii

Note:  I originally wrote this article for the General Aviation Council of Hawaii Fall 2012 newsletter. Had to make a couple tweaks to the dual-polarization section, since upgrades at all Hawaii radars were completed in 2013. –JB

Radar image of a supercell thunderstorm from March 9th, 2012. This storm brought large hail to Kailua and Kaneohe, including a record-setting hailstone that measured 4.25 inches long, and generated a tornado that damaged homes in Lanikai.

Radar image of a supercell thunderstorm from March 9th, 2012. This storm brought large hail to Kailua and Kaneohe, including a record-setting hailstone that measured 4.25 inches long, and generated a tornado that damaged homes in Lanikai.

Weather radars are one of the most effective tools for detecting rainfall. This article includes information about weather radar in general, and about the specific radar sites we have in Hawaii.

Weather Radar Basics

The radar transmits an electromagnetic pulse, which reflects off objects in the atmosphere. (The objects could be anything, not just precipitation; one common problem around Hawaii is with sea spray being detected during windy days.) A small fraction of the energy is returned to the radar. The radar measures how much energy is reflected back, and how long it took to return. The energy is converted into reflectivity, and the time is converted into distance from the radar. Larger or more numerous objects return more energy, and result in a higher reflectivity.

Continue reading

Mountain Turbulence

Note:  I originally wrote this article for the General Aviation Council of Hawaii Winter 2012 newsletter.  Hopefully you will find it interesting and educational as well. –JB

Two common types of turbulence associated with mountains are mechanical turbulence and mountain waves. Mechanical turbulence is a result of an obstruction to the wind flow. Obstructions can range in size from trees and buildings to rough terrain and mountains. The degree of the turbulence depends on the strength of the wind speed and the size and shape of the obstruction. The stronger the wind or the rougher the terrain, the stronger the turbulence will be.

Mechanical turbulence usually occurs within 20 miles of the mountain, and is located at an altitude near or below the height of the terrain. As a rough estimate, low-level winds of 20 knots may lead to light turbulence, winds of 25 knots may lead to moderate turbulence, and winds greater than 30 knots may lead to severe turbulence.

Trapped lee waves downwind of a mountain range.  Image courtesy of COMET/UCAR

Trapped lee waves downwind of a mountain range. Image courtesy of COMET/UCAR

For mountain waves, there are two main types: trapped lee waves and vertically propagating mountain waves. Mountain waves are a type of gravity wave, meaning that they are forced to oscillate because of gravity. (Waves on the ocean are another type of gravity wave.) When strong winds blow across a mountain, the air is forced upward by the terrain. If the atmosphere over the mountain is stable (that is, if the temperature of the air increases with height, which is typically the case in Hawaii due to the trade wind inversion), then the air that is forced upward by the mountain will be more dense than the air around it. The air will sink back toward the ground, where it will begin an up and down oscillation. These oscillations can continue for over 50 miles downstream of the mountain, and are known as trapped lee waves.

Visible satellite image of wave clouds over and downstream of Oahu under strong southwest winds.

Visible satellite image of wave clouds over and downstream of Oahu under strong southwest winds.

If the atmosphere has enough moisture, the trapped lee waves may be visible as wave clouds. Wave clouds form near the crests of the trapped lee waves, and appear as distinct lines that are oriented parallel to the terrain and perpendicular to the wind. While the clouds may appear to remain stationary, the wind blowing through them is actually quite strong.

Trapped lee waves tend to form when the low-level flow is perpendicular to the mountain range, there is an inversion located near the top of the ridge, and ridge-top winds are 25 knots or greater. Wave clouds typically form at an altitude within a few thousand feet of the ridge top. Turbulence is most often encountered below the crests of the mountain waves, or, in other words, below the wave clouds if they are present.

Trapped lee waves are the most common type of mountain wave in Hawaii, because of the persistence of the trade wind inversion. However, if the atmosphere is unstable, vertically propagating mountain waves may form. In this situation, there is no inversion for the mountain waves to reflect off of to begin the oscillation. Instead, the mountain waves will spread upwards through the atmosphere, and also tilt upstream in the direction from which the wind is blowing. Vertically propagating mountain waves may extend through the troposphere and even into the stratosphere.

The presence of vertically propagating mountain waves does not necessarily mean that there will be turbulence. If the amplitude of the waves becomes large enough, they become unstable and break–just like ocean waves–which leads to turbulence. If the waves don’t break, an aircraft may encounter significant wave action, but not the severe to extreme turbulence that may be encountered within breaking mountain waves.

The conditions necessary for mechanical turbulence and mountain waves are similar: strong winds blowing across a mountain. Whether the cause is mechanical turbulence or mountain waves, the results can be the same: turbulence. As a basic rule of thumb, be alert for turbulence if low-level winds are 25 knots or greater. The National Weather Service will issue an AIRMET if moderate turbulence is expected, and will issue a SIGMET if severe turbulence is expected. When an AIRMET or SIGMET is in effect, the forecaster will also provide the reasoning behind those products in the Area Forecast Discussion. These products (and others) are available through the aviation page of the WFO Honolulu website.

John Bravender
Aviation Program Manager
National Weather Service Honolulu

La Niña and Hawaii Weather

Note:  I originally wrote this article for the General Aviation Council of Hawaii Fall 2011 newsletter. The seasonal outlook no longer applies, but typical impacts during La Niña remain consistent.  Hopefully you will still find it interesting and useful. –JB

The fall 2011 forecast [view the latest forecast] from the National Weather Service Climate Prediction Center indicates that La Niña conditions have returned, and are expected to gradually strengthen and continue into the winter of 2011-2012. What is La Niña and what does it mean for weather in Hawaii this wet season?

What is La Niña?

Observed sea surface temperature and SST anomaly from late September, 2011.

Observed sea surface temperature and SST anomaly from late September, 2011.

La Niña is the cold phase of the El Niño/Southern Oscillation (ENSO) cycle. It is identified by colder than normal water along the equatorial Pacific. It is the opposite of El Niño, which is identified by warmer than normal water along the equatorial Pacific. ENSO episodes typically last 9-12 months, and reach their peak strength during late winter (from December to April). The episodes typically occur every 2-7 years. However, it is not uncommon for a strong La Niña (such as the one that occurred during the winter of 2010-2011) to be followed the next year by a weak La Niña.

The changes in the ocean temperature have wide-ranging impacts on atmospheric circulations and weather patterns.

Weather Impacts to Hawaii

During La Niña years, large scale flow across the eastern North Pacific tends to be more amplified. Instead of a zonal (west to east) jet stream, it can develop more of a north/south component. This amplified pattern can cause storm systems to track farther south than normal. In addition, a persistent upper level high is also common over the Gulf of Alaska. This “blocking” pattern can cause low pressure systems to linger over one area for a prolonged period of time.

At the local scale, Hawaii tends to see above normal rainfall and more frequent storm systems during La Niña years. The winter of 2010-2011 occurred during a strong La Niña. There were many heavy rain episodes, including a number of widespread thunderstorm events. The storms continued into early June, well beyond the typical wet season.

There are a number of aviation impacts that are related to the increased number of winter storms that affect Hawaii during La Niña years. Thunderstorms are more frequent near the islands, with their attendant hazards: severe or greater turbulence, severe icing, wind shear, and IFR conditions. Mountain obscurations are more common and can be more persistent. Widespread icing is possible, and icing conditions can be encountered at lower altitudes.

While these hazards occur regardless of the ENSO cycle, they tend to happen more frequently during La Niña years. The current long-range outlook for Hawaii calls for above normal rainfall and above normal temperatures for the second half of the winter. As the wet season progresses, remember to stay up to date with the latest warnings, advisories, and forecasts from your National Weather Service. If you have questions about any NWS forecast products, you can call the office at 808-973-5286. (Just remember that while we can answer your questions, we can’t provide flight briefings.)

John Bravender
Aviation Program Manager
National Weather Service Honolulu

A Quick Introduction to Your National Weather Service Office

Note:  This was my first article for the General Aviation Council of Hawaii newsletter.  It was first published in Summer 2011. –JB

Map of the WFO Honolulu aviation area of responsibility. Aviation forecasts include SIGMETs, AIRMETs, TAFs, and ROFORs.

Map of the WFO Honolulu aviation area of responsibility. Aviation forecasts include SIGMETs, AIRMETs, TAFs, and ROFORs.

Located on the UH-Manoa campus, the National Weather Service (NWS) Weather Forecast Office (WFO) in Honolulu provides aviation forecasts and warnings for over 8 million square miles of the central and western Pacific. The office has 39 employees, including meteorologists, technicians, and support staff. There are at least four meteorologists on duty 24 hours a day, with one person dedicated to aviation forecasts.

WFO Honolulu is one of three Meteorological Watch Offices (MWOs) in the United States, along with the Aviation Weather Center in Kansas City, MO, and the Alaska Aviation Weather Unit in Anchorage, AK. MWOs are designated by the International Civil Aviation Organization to maintain a continuous watch over weather conditions that affect flight operations, and to issue necessary warnings and forecasts for the aviation community.

As a forecast office and MWO, WFO Honolulu is responsible for the following aviation products:

  • SIGMETs for the Oakland Oceanic FIR south of 30°N and west of 140°W
  • AIRMETs for the Main Hawaiian Islands (out to 40 nm from shore)
  • Area Forecasts for the Main Hawaiian Islands (out to 40 nm from shore)
  • TAFs for the following 9 airports in Hawaii, plus Midway (MDY) and Pago Pago, AS (PPG)
    • LIH, JRF, HNL, MKK, LNY, JHM, OGG, KOA, ITO
  • Route Forecasts from SFO/SBA to HNL for low-level flights
  • Forecast Discussions for the Main Hawaiian Islands describing the meteorological thinking behind the forecast
  • Winds/Temperatures Aloft (automated from computer model data)

If you have question about any of these forecast products, you can call the office at 808-973-5286. (Just remember that while we can answer your questions, we can’t provide flight briefings.)

This has been just a brief overview of what we do at WFO Honolulu. In addition, I would like to encourage you to provide feedback to us. The PIREPs we receive in real time are extremely helpful to our aviation forecasters. In addition, we always welcome constructive comments and suggestions for improvement. Please feel free to contact me either through the phone number above, or the email address below.

John Bravender
Aviation Program Manager
National Weather Service Honolulu

A Collaborative Turbulence Observation Project

Note:  I originally wrote this article for the General Aviation Council of Hawaii January 2015 newsletter, hence the aviation focus.  Hopefully you will find it enjoyable and educational as well. –JB

As the forecaster who oversees the aviation weather program at WFO Honolulu, I’m always looking for ways to work with the aviation community to improve the services we provide. Both meteorologists and pilots can benefit from more detailed observations of turbulence around the islands. In this article I’ll outline an idea for a “citizen science” project that could do just that.

Visible satellite image showing wave clouds over and to the southwest of Oahu.  7am HST, August 18th, 2012.

Visible satellite image showing wave clouds over and to the southwest of Oahu. 7am HST, August 18th, 2012.

One thing that forecasters struggle with is how far away from the terrain that turbulence occurs. As mentioned in the January 2012 article on mountain waves, we use general proxies to determine when turbulence may occur—for example, low-level winds of 25 knots or greater—and correlate these with pilot reports. PIREPs are very important in this process, because they are the only way we have to know what is actually occurring. However, PIREPs are usually provided at just one point and the pilot is looking to get out of the turbulence as fast as possible, not trying to see how far it extends. Satellite imagery can help in some instances, but the cloud features that indicate turbulence—such as wave clouds extending downstream of the mountains, as seen over and to the southwest of Oahu in this image—aren’t always present. We need some way to get more information about where turbulence occurs.

Aircraft reporting eddy dissipation rate (EDR) from December 24th, 2014.

Aircraft reporting eddy dissipation rate (EDR) from December 24th, 2014.

Increasingly, large aircraft are being equipped with instrument packages that include accelerometers. An algorithm was developed that uses the vertical acceleration of the aircraft to estimate a parameter called eddy dissipation rate (EDR). It provides an objective measure of how much turbulent motion the aircraft encounters, and is continuous along the flight track. The adjacent image shows a number of flights reporting EDR from December 24th, 2014. The blue and orange dots indicate no turbulence, magenta indicate light to moderate turbulence, and red (like the one circled) indicate moderate to severe turbulence. However, there is no need to resort to expensive new equipment to get this type of measurement. Many of us have smart phones that can do the same thing.

Map of bicycle "turbulence" measured using a smartphone accelerometer.

Map of bicycle “turbulence” measured using a smartphone accelerometer.

The adjacent map shows a short bicycle ride that I took using the accelerometer in my phone to measure the bumpiness of the route. The larger orange and red circles are bumpier conditions: potholes in the lower right, speed bump in the upper left, and embedded reflectors in the lower left. (The data were plotted using GPS Visualizer). Something similar could be used in aircraft as well. By using an app that logs GPS, altitude, and acceleration, we can get a measure of how much turbulence occurred and where it occurred during the course of a flight. We also need to normalize the data. (What does “20 [units]” mean after all?) For the commercial aircraft mentioned above, they use a mathematical model to determine how the aircraft responds to turbulence. However, this can also be done in a subjective manner. The pilot can take the flight track and highlight a few points—it was smooth in these areas, I had light turbulence here and here, and at this point I encountered moderate. With a few reference points, we can apply those values to the rest of the route and get turbulence observations for the entire flight track.

This would be an immense help to forecasters, since it will show us in much greater detail where turbulence occurs and how it changes across the area. The information can also be made available to other pilots (as an aggregate product, stripped of identifying information) in near real time as a flight planning tool, as well as part of a climatology that can help new pilots learn of dangerous areas to avoid.

However, this is just an idea right now—which is why this article includes a bicycle track instead of a flight track. Eventually we’ll be looking for pilots who want to participate by measuring data as they fly. In the meantime I would love to hear comments or suggestions. Is this something that is feasible? Is it something that you’d be willing to do on a regular basis? As I mentioned, this is just the starting point of the process, and now is one of the best times to share your insight and expertise to help make it successful.

Thanks for your help!

John Bravender
Aviation Program Manager
National Weather Service Honolulu

Aircraft Reconnaissance and Tropical Cyclones

Note:  I originally wrote this article for the General Aviation Council of Hawaii Fall 2014 newsletter, hence the aviation focus.  Hopefully you will find it enjoyable and educational as well. –JB

Visible satellite image showing Hurricanes Iselle and Julio in the Central Pacific and Typhoon Genevieve just west of the dateline. Image taken 2pm HST Thursday, August 7, 2014.

Visible satellite image showing Hurricanes Iselle and Julio in the Central Pacific and Typhoon Genevieve just west of the dateline. Image taken 2pm HST Thursday, August 7, 2014.

The 2014 Central Pacific hurricane season started off with a very active period from late July through mid August. As Genevieve sputtered south of the state, it became clear that Iselle and Julio had the potential to be a significant threat to Hawaii. During most situations, forecasters at the Central Pacific Hurricane Center in Honolulu rely on satellite imagery to estimate the position and intensity of tropical cyclones. However, when there is the threat of a cyclone making landfall, NOAA and the U.S. Air Force Reserve will dispatch aircraft to fly into the cyclone. Instead of having to infer the intensity of a system from satellite, aircraft reconnaissance provides actual measurements.

This article will take a look at the organizations and aircraft that fly into tropical cyclones, the types of equipment that they use to help meteorologists make more accurate and timely forecasts, and how it all came together during the first tropical storm landfall in 22 years in Hawaii.

Organizations and Aircraft

There were two organizations performing reconnaissance on Hurricanes Iselle and Julio: the U.S. Air Force Reserve’s 53rd Weather Reconnaissance Squadron and the NOAA Hurricane Hunters. The 53rd WRS flies WC-130J Super Hercules turboprops, and NOAA flies two WP-3D Orion turboprops and a Gulfstream IV-SP turbofan jet. The turboprop aircraft are used for flying through a hurricane, while the jet is used to sample the environment outside of a hurricane.

WC-130J Super Hercules flown by the 53rd Weather Reconnaissance Squadron "Hurricane Hunters"

WC-130J Super Hercules flown by the 53rd Weather Reconnaissance Squadron “Hurricane Hunters”

The USAFR 53rd WRS “Hurricane Hunters” are based out of Keesler Air Force Base in Biloxi, MS. They operate 10 WC-130Js, and deployed four of their planes to Joint Base Pearl Harbor-Hickam. They flew tracks through both Iselle and Julio, identifying the center position and strongest winds/lowest central pressure as well as the structure and size of the wind field. Typical missions last about 10 hours, and normally include center position fixes every few hours. The data collected by the aircraft is transmitted in bursts as it is collected, and is available to hurricane specialists—and the general public—in near-real time.

Gulfstream IV-SP and WP-3D Orion flown by the NOAA Hurricane Hunters

Gulfstream IV-SP and WP-3D Orion flown by the NOAA Hurricane Hunters

The NOAA Hurricane Hunters are managed by the NOAA Aircraft Operations Center, based out of MacDill Air Force Base in Tampa, FL. They operate two types of aircraft for hurricane research: two WP-3Ds (nicknamed “Kermit” and “Miss Piggy”) and one Gulfstream IV-SP (a relatively recent addition, nicknamed “Gonzo”). In terms of Iselle and Julio, it was the Gulfstream that provided reconnaissance. “Gonzo” samples the environment around a cyclone, and the data it collects is included in the computer models that hurricane specialists use to forecast the track and intensity. (Research has shown that proper sampling of the atmosphere around a cyclone can improve the track forecast by up to 25% within the first two days.)

Weather Observations

Flight level winds (10,000 feet) and vortex center fixes from mission 7 into Iselle Thursday evening, August 7, 2014. This was the mission that determined Iselle had weakened to a tropical storm just before landfall.

Flight level winds (10,000 feet) and vortex center fixes from mission #7 into Iselle Thursday evening, August 7, 2014. This was the mission that determined Iselle had weakened to a tropical storm just before landfall.

Weather instruments on the aircraft provide continuous observations of winds, temperature, and dew point at flight level. Using an empirically derived reduction value, the flight level wind can be used to approximate the surface wind. The surface wind is about 90% of the flight level wind at 10,000 feet, or about 80% of the flight level wind at 5,000 feet. (It might sound strange to reduce the lower altitude winds more, but because of the unique structure of a hurricane, the strongest winds are located at about 1,500 feet.)

To actually measure the wind at the surface, the aircraft carry instrument packages called dropsondes. These instruments are similar to those used with weather balloons. However, instead of being attached to a balloon and measure the atmosphere from the surface upward, they are attached to a parachute and measure the atmosphere from the plane downward. They measure temperature, dew point, wind speed and direction, and pressure. Multiple sondes are dropped throughout a mission, with at least two released during a center fix (one as the flight passes in through the eye wall, and one as it passes out through the other side of the eye wall).

One of the newer additions to the observation equipment is the Stepped Frequency Microwave Radiometer (SFMR). Whereas a dropsonde can only measure the wind at one point, the SFMR provides a continuous measure of winds at the ocean surface below the plane. It detects naturally emitted microwave radiation from sea foam, which it uses to calculate wind speed and direction. It can also estimate rainfall intensity.

A Doppler radar in the nose of the aircraft also provides information on the reflectivity and velocity structure of the cyclone. This information lets forecasters know how well organized a cyclone is and how it’s changing over time. For example, ragged looking rain bands that start to become more organized indicate a cyclone is strengthening. The radar observations are also used to improve the initial starting point for the computer simulations. (The better picture you have of the current state of the atmosphere, the better the computer model will perform.)

Flight Tracks

Flight track for mission #3 into Hurricane Iselle from August 6, 2014

Flight track from mission #3 into Hurricane Iselle from Wednesday, August 6, 2014. White arrows show the direction of flight.

The USAFR Hurricane Hunters fly through the center of the cyclone to measure the strongest winds and lowest pressure, as well as get an accurate center position. They fly what is known as an alpha pattern (based on its resemblance to the Greek letter α). Each leg is normally just over 100 nautical miles long, and is flown on diagonal tracks (NW-SE, SW-NE). This image shows the flight track from mission #3 into Hurricane Iselle from August 6th, 2014. The plane entered the cyclone from the northwest, and proceeded to complete two alpha patterns, providing four separate eye penetrations. These are denoted by the white circles; the number is the minimum central pressure that was measured in the eye. The times of the four fixes were 17:45 UTC (7:45 AM HST), 19:44 UTC, 21:21 UTC, and 23:01 UTC, or roughly an hour and a half to two hours between fixes. During this mission, they also released seven dropsondes during the eye penetrations.

Flight track for mission #4 into Hurricane Iselle from Wednesdy August 6, 2014

Flight track from mission #4 into Hurricane Iselle from Wednesday, August 6, 2014. The flight departed northeast from Oahu, proceeded clockwise around the hurricane, and returned south of the Big Island.

The NOAA Hurricane Hunters WP-3D aircraft fly through cyclone centers, similar to the WC-130Js. However, the Gulfstream is used to sample the environment around a cyclone. Instead of crisscrossing through the eye of a hurricane, they fly around it, releasing dropsondes around the periphery. They also release a number of sondes ahead of the cyclone, in order to get a better picture of the environment that it will be moving into. This image shows the flight track from mission #4 into Hurricane Iselle from August 6th, 2014. During this flight they made 32 drops. The first one was northeast of Maui at 18:11 UTC (8:11 AM HST), and the last one was just south of the Big Island at 00:22 UTC.

Landfall

Iselle made landfall at about 2:30 AM HST Friday morning, August 8th, 2014, along the Kau coast of the Big Island, approximately five miles east of Pahala. Earlier in the evening, aircraft reconnaissance only measured maximum flight level winds of 60 knots and maximum surface winds of 57 knots, and the aircraft radar saw no discernable eye wall. Since a hurricane is defined as having winds of 64 knots or greater, Iselle had weakened to a tropical storm just before landfall. Even with the weakening trend, there was significant wind damage across Puna District on the Big Island and across upcountry areas on Maui. Heavy rain also caused flooding problems, which were greatest across Kau District where the center of the circulation moved onshore. Several rain gages over the upper slopes of South Hilo and Puna Districts received over 10 inches of rain.

Iselle weakened rapidly after interacting with the terrain of the Big Island. A low-level circulation center redeveloped west of the Big Island early Friday morning, but the system never regained its earlier organization. Tropical storm warnings for the smaller islands were cancelled Friday afternoon, and all tropical bulletins were discontinued Friday night.

Just When You Think It’s Over

In the days leading up to landfall, it looked like Iselle and Julio could pack a one-two punch. After going 22 years without a landfall, it was possible that we could have two in just a couple days. However, by Sunday, August 10th, 2014, it became clear that Hurricane Julio would stay northeast of the state and not impact Hawaii or other U.S. interests. Therefore, the decision was made to end reconnaissance, and mission #7 was deemed complete after only two of the four scheduled fixes.

As the plane began to return to Honolulu, the U.S. Coast Guard received a distress call from a sailboat near the center of the hurricane. The boat had lost its engine and was taking on water. The location they provided was just northeast of the eye, which, for a northwestward moving hurricane, is the area with the most dangerous seas. It was also an area where the reconnaissance aircraft had just measured 55 knot winds at the surface.

Flight track from mission #7 into Hurricane Julio from Sunday, August 10, 2014. White arrows show the direction of flight, and the white circle highlights the search area.

Flight track from mission #7 into Hurricane Julio from Sunday, August 10, 2014. White arrows show the direction of flight, and the white circle highlights the search area.

Communications between the sailboat and the Coast Guard Joint Rescue Coordination Center were spotty at best, and it would take time for the Coast Guard HC-130H from Barbers Point to reach their last reported position and begin a search. The WC-130J, which was just leaving the area of the hurricane, turned around and began the search. Because they had planned for a 10 hour mission that ended up being truncated halfway through, they had fuel to spare. They established radio communication with the boat, and eventually were able to spot them after dropping to a low altitude. Once the boat’s location was confirmed, the mission was turned over to the Coast Guard aircraft, and the Hurricane Hunter returned to Honolulu.

Luckily the crew of the sailboat made it home safely. Although they weren’t able to reach the supplies that the Coast Guard dropped for them, the Matson containership Manukai reached them that night, and brought them aboard first thing Monday morning.

Epilogue and Feedback

Hopefully you enjoyed this article. There’s one more piece of information you might like, which will further allow you to play along at home during the next tropical cyclone. The maps of the reconnaissance data were created using Google Earth, with KML files from Tropical Globe.  These files are updated in near-real time as reconnaissance is received.  An archive of past reconnaissance observations in available for Iselle and Julio.

Also, we are always looking for feedback. How did Hurricane Iselle affect your aviation operations across the state? Did the products issued by the Central Pacific Hurricane Center and WFO Honolulu meet your needs? Was there anything in particular we did that you really liked? Really didn’t like? What would you like to see in the future with respect to aviation weather services?

Thanks for any feedback you care to provide!

John Bravender
Aviation Program Manager
NOAA/National Weather Service
Central Pacific Hurricane Center
Weather Forecast Office Honolulu