Pacific ENSO Update - November 1995 - Vol.1 No.4


Weather patterns in the tropical Pacific have been influenced by a trend toward "cold event" conditions, following the end of the most recent (1994-95) ENSO warm event. Cold events, also called "La Niña", usually occur less frequently than ENSO warm events. But when they do occur, it is nearly always just after a warm event. The most recent example of this sequence is the 1988-89 La Niña, which followed the ENSO warm event of 1987-88. Consistent with La Niña conditions, easterly winds have been occurring throughout the equatorial Pacific at slightly-above normal strength in recent months. These winds have accompanied a cooling trend in sea surface temperatures (SSTs) along the equator, extending from the east to west (currently to about 160 degrees E longitude, between the Solomon Islands and Pohnpei). At the present time, these conditions are not too far beyond their normal ranges. However, the behavior of large-scale weather patterns, including decreased tropical cyclone activity in the central and western tropical Pacific, late arrival of the monsoon trough, and below-normal rainfall throughout much of Micronesia in spring and summer, has been consistent with a post-ENSO trend toward La Niña conditions.

New issues of the "ENSO Advisory", describing the extent of the current "cold phase" conditions and the outlook for normal-to-cool conditions in the coming months, have recently been produced by NOAA and are discussed in this issue of Pacific ENSO Update. Local variability summaries for individual island areas, and descriptions of new experimental forecasts produced by NOAA for Pacific island rainfall, are also included in this issue, along with some information about long-term forecasts for various ENSO conditions from experimental climate models.


Earlier issues of Pacific ENSO Update have included descriptions of the "ENSO Advisory". An "ENSO Advisory" is a special document produced by NOAA's National Weather Service when ENSO conditions are occurring in the tropical Pacific. The last sequence of ENSO Advisories ended in April 1995, when an end to the 1994-95 ENSO warm event was apparent. A new issue, dated October 18, 1995, describes some trends toward ENSO "cold phase", or La Niña, conditions. The contents of this issue of the ENSO Advisory are shown in Appendix I. Future monthly issues of the ENSO Advisory are expected to continue reporting on the development of the La Niña cold phase conditions.


La Niña conditions are usually the opposite of those associated with El Niño-Southern Oscillation (ENSO) warm events. During La Niña, stronger-than-normal easterly (trade) winds occur throughout much of the tropical Pacific, in association with higher-than-normal atmospheric surface pressure at Tahiti and lower-than-normal pressure at Darwin, Australia - a condition indicated by positive values of the Southern Oscillation Index (SOI). These strong winds push greater amounts of warm surface waters into the far western Pacific. This allows colder, deeper water to well up along the equator in the central and eastern tropical Pacific and the coast of South America. Greater amounts of cloudiness and rain occur over the warmer waters in the far western Pacific. Areas to the east of there can receive less rain. These conditions can also reduce tropical storm and cyclone development in eastern areas.

As described in the ENSO Advisory (Appendix I), the current intensity of these cold conditions is mixed, with cooler sea surface temperatures (SSTs) but near-normal wind patterns. In addition, the development of these conditions has not been as fast or as consistent as has been seen in earlier events, like the La Niña of 1988-89. At present, these conditions are not expected to change much within the next few months, but future issues of the ENSO Advisory and Pacific ENSO Update will report on changes as they develop.

Special Section: CLIMATE MODELS

In contrast to the ENSO Advisory, which is produced only on certain occasions, NOAA's National Weather Service regularly produces two much more detailed publications: the Climate Diagnostics Bulletin, produced monthly, and the Experimental Long-lead Forecast Bulletin (ELFB), produced quarterly. These two publications provide many detailed kinds of information related to climate around the world, including a large amount of ENSO-related information. The Climate Diagnostics Bulletin primarily reports on detailed observations of recent atmospheric and oceanic conditions, and on operational climate forecasts. The ELFB gives the results of several experimental forecasts and predictions on features of the ENSO climate cycle and related weather patterns. These experimental forecasts are computer-based "models" which use one or both of the following techniques for climate prediction:

A visual representation of these techniques is shown in the figures below...
Source: "El Niño and Climate Prediction" -
NOAA's "Reports to the Nation" series

STATISTICAL CLIMATE MODELS take advantage of observed relationships between different features of the climate system. In the above diagram, an apparent relationship is seen between: 1) sea surface temperatures near South America [top graph], 2) atmospheric pressure at Darwin, Australia [middle graph], and 3) annual rainfall at Christmas Island in the central equatorial Pacific [bottom graph]. Statistical models require long-term historical records to detect these relationships, then use them to make predictions about future climate conditions.

DYNAMICAL CLIMATE MODELS attempt to approximate the natural processes which occur to produce winds, currents, sea surface temperatures, and rainfall. Unlike statistical models, dynamical models use cause-and-effect relationships and can predict conditions which have not been observed before at locations which may not have long-term historical records.

Both these techniques have their advantages and disadvantages:

Statistical models are comparatively simple to design, can detect and take advantage of apparent relationships, and perform very well for predicting conditions within a range of what has been experienced before. But, when record conditions occur (e.g. conditions that lead to record floods, droughts, and their serious impacts), statistical models are less able to produce a reliable forecast, because those conditions can fall outside the range of what has been observed before.

Dynamical models are designed to behave in the same way that nature does, and can predict wide ranges of conditions, including record conditions which have not been observed before. But, they are limited by our understanding of the actual physical processes and our ability to make mathematical approximations of them. They are also very sensitive to the quality, quantity, and application of observed data within the model system. Small errors in input data can lead to large errors in output. Or, inadequate levels of "resolution" - the scale at which the model attempts to represent the natural climate system - can lead to incorrect predictions.

Both types of models have measurable skill, which is called "useful" if it exceeds our ability to make reasonable predictions using "climatology", or another method called "persistence":

CLIMATOLOGY can be thought of as a "history" of conditions in the atmosphere (e.g. weather) and the ocean. It provides a combined source of information about average conditions and their known ranges of variability, based on data and observations. As such, climatologies can be defined for any feature of the climate system, and at any time scale, for which data and observations are available. Traditionally, climatologies of temperature and rainfall are based on 30-year periods. However, different 30-year periods can yield different climatologies. In a sense, climatology is the most basic sort of "prediction" that can be applied to expected future conditions. But climatology alone has no skill at prediction of one condition over another, beyond random chance. Instead, climatology provides an essential background for other kinds of prediction techniques, which attempt to forecast future conditions with greater certainty than climatology alone can provide.

PERSISTENCE, in simple terms, is a prediction technique which is essentially based on the continuation of trends -- the expectation that "tomorrow will be like today", "this month should be like last month", or "this winter should be like last winter". This age-old technique for predicting weather and climate conditions is actually quite good at lead times of up to a few days for things like rainfall, to weeks and months for things like wind patterns and sea surface temperatures. But at lead times of months to a year or more, persistence loses its usable skill over climatology. It is on these lead times that statistical and dynamical models can beat persistence, and provide predictions of greater certainty than is possible with climatology alone, especially for the climatic conditions related to the ENSO cycle.


The latest issue of the Experimental Long-lead Forecast Bulletin (ELFB - see "special section" above) gives the results of several experimental models for forecasting ENSO conditions. A summary from the September 1995 ELFB appears below:


For ENSO Condition:

DYNAMICAL METHODS:  The Scripps/MPI HCM predicts -1 deg.C SSTA 
in eastern Pacific fall to early winter '95-96, followed by 
strong warming summer-fall '96.  The NCEP coupled model calls 
for -1 deg.C SSTA east-central Pacific through late winter '95-96.
The Cane-Zebiak model predicts somewhat below-normal Nino 3 SST
through spring '96, followed by marked warming summer-fall.  
The Australian BMRC low order coupled model forecasts current 
cool Nino 3 northern fall '95 to become warm spring '96, increasing
to 1.5 deg.C fall '96.  The Oxford coupled model calls for current
normal to coolish SST to continue through winter '95-96, with
slight further cooling spring-summer.  The COLA coupled model
predicts near-normal SST conditions to winter '95-96, hint of 
very slight warming in mid-96.

STATISTICAL METHODS:  The Penland inverse modeling predicts
slightly below normal east-central Pacific SST fall to winter 
95-96, returning toward normal by summer.  The Australian BMRC
non-linear analogue predicts near-normal SOI through this northern
fall.  The UCLA/JPL SSA-MEM predicts near normal SST through 
mid-'96 and near normal SOI over the next 4 years.  The NCEP/CPC's
CCA predicts slightly below normal Nino 3.4 (120W - 170W) SST
fall '96, returning to near normal by late winter, through summer
'96.  NOAA's constructed analog predicts near-normal Nino 3.4
through winter '95-96, becoming slightly warm spring through 
fall '96.
In general, the dynamical models are in agreement for predicting near-normal conditions to continue through the early part of 1996. However, three of six models are predicting the development of an ENSO warm event, possibly of strong magnitude, beginning in spring-summer and persisting through fall-winter of 1996/97. The remaining dynamical models predict either near-normal conditions or slightly cool conditions over various portions of this period.

Nearly all of the statistical models are predicting close-to-normal conditions through the first half of 1996, with one of five models predicting a slight warming trend that would be consistent with a 1996/97 ENSO warm event, but at a magnitude that is closer to normal conditions.

Given this range of results in the models, and the limits of their skill at longer lead times, the following LOCAL VARIABILITY SUMMARIES are based on an expectation that conditions will be near-normal through 1996. Any changes in trends, and their impact on the outlooks in these LOCAL VARIABILITY SUMMARIES, will be reported in future issues of Pacific ENSO Update.


As noted in each issue of Pacific ENSO Update, the following summaries of expected local climate variability for the various island areas indicated are not based on official forecasts (unless otherwise noted). Further information is available from your local National Weather Service office, or from the various sources listed in the acknowledgements section.

HAWAII: Weather in the Hawaiian Islands has turned to normal conditions in the transition between the warm, dry (Kau) season and the cool, wet (Hoo'ilo) season. Recently, two episodes of unusually strong troughing in the mid-troposphere produced strong thunderstorms and heavy rains on Kauai, Oahu and Hawaii. Heavy thunderstorm rains drenched leeward parts of Oahu and Kauai, providing the first significant rains of the year. Shearing wind fronts have also brought early season rainfall to the the northern islands.

In sharp contrast to the Atlantic, the 1995 hurricane season in the eastern and central tropical Pacific has been the quietest since 1979. Only one decaying system (remnant of tropical storm Barbara) entered the region west of 140 degrees W longitude.

The most recent issue of the Long-Lead Outlook for the Hawaiian Islands, issued by NOAA's Climate Prediction Center, is included in this bulletin. Outlooks for 3-month periods throughout the coming year are for rainfall and temperature conditions close to normal.
- sources: U.H. Dep't of Meteorology and NOAA-CPC

AMERICAN SAMOA: Dry season rainfall at Pago Pago from July-September was near its long-term average. With the trend in La Niña conditions described in the ENSO Advisory (see Appendix I), rainfall in American Samoa and nearby regions is expected to be about 10-15% below the long-term average for the remainder of 1995. By comparison, rainy season rainfall in this area is about 20-25% heavier during ENSO years.

The recent development of cool conditions should reduce the chance of tropical cyclone and storm impacts in the region near Samoa. Hurricane and tropical storm activity for the coming 1995-96 season should be primarily in western areas between Fiji and Australia. Activity in areas east of the date line, including Samoa, is expected to be below average.
- source: UOG-WERI

GUAM/CNMI: Rainfall increased substantially in late August, September, and October, and has been above normal on Guam and in the CNMI, despite the low tropical cyclone activity during 1995. The rainfall has been more than sufficient to sustain lush vegetation. On Guam, the rainfall has completely filled the Fena reservoir, overflowing the spillway.

The meteorological conditions that usually produce heavier summertime rains in Guam and the CMNI were about two months late in setting up for 1995. This extended drier-than-normal conditions into late August. The normal monsoon pattern did not become well established over the region, but two short episodes of moderate monsoon conditions, some weak storm activity, and nearly daily convective showers in September and October brought total rainfall in at about five inches above the long-term average for those months on Guam and in the CNMI. Therefore, total wet season rainfall for Guam and the CNMI should be just about normal, unless November and December are extremely dry, which does not seem likely given the slow rate of change in current conditions.

New results from experimental work on climate prediction in the Pacific islands are indicating the possibility of a below-normal trend in rainfall for northern Micronesia and the Mariana Islands over the next several months (see the section "New Experimental Forecasts for Pacific Island Rainfall, pp. 8-9). The following expectations of monthly rainfall totals reflect this trend for the months of November through March. Beginning in April, the outlook for a trend toward normal conditions in the spring and summer of 1996 is reflected in the expectation for slightly higher-than-average rainfall. Since the ENSO and La Niña are associated with rainfall below the long-term average for the region, "normal" conditions are expected to bring rainfall in amounts above the long-term average.

Per their request, the following expectations of rainfall totals are provided for the use of water resource managers on Guam:

	  Nov 95:	  	100% of long-term average
	  Dec 95 - Apr 96:	 90% of long-term average
	  Apr 96 - Nov 96:	120% of long-term average
Tropical cyclone activity in the western tropical North Pacific during 1995 was at a near-record low, and centered further west than usual. This is consistent with the trend toward La Niña cold conditions that developed over recent months. With the expectation of a gradual return to normal conditions, typhoon activity should return to normal with respect to when and where they occur. A chance of some activity remains for the current season through December 1995. Thereafter, some possibility of late-spring and early-summer activity in 1996 exists for central and western Micronesia, including Guam and the CNMI, with greatest risks for the region resuming at their normal time from mid-September through Mid-December 1996.
- source: UOG-WERI

MICRONESIA: In this issue of Pacific ENSO Update, the local variability summaries for the FSM have been changed to reflect each state separately:

YAP STATE: Rainfall in Yap State is strongly influenced by ENSO and La Niña conditions. During the spring and summer of 1995, rainfall in northern Yap State islands was around 50% below the long-term average. The southern islands of Yap State also had extended periods of below-average rainfall during this time, but not to the same degree as was seen in the northern islands. By September, southwest monsoon activity brought rainfall back to most islands in near-average amounts. October rainfall on Yap Island and Ulithi was over 40% above the long-term average for the month, but this was primarily due to the rainfall from Tropical Storm Zack and Typhoon Angela. For the coming months, rainfall is expected to be near-average across the State until about April 1996, then about 20% above average for the rest of the year. With a return to normal climate conditions in 1996, above-average rain is expected in Yap State, since ENSO and La Niña years are drier than average. Tropical cyclone activity for Yap State is also expected to return to near-normal. This means that Yap, Ulithi, and Fais can expect their greatest threats from September through November of 1996, with some possibility of threats in May and June. The southern islands can expect their greatest threats slightly later, from October through mid-December 1996, with lesser threats during April and May.

CHUUK STATE: While September rainfall was about 20% above the long-term average for Weno Island, October rainfall fell to values slightly below normal. Rainfall for all Chuuk State islands is expected to be near average through March 1996, and about 20% wetter than average after March 1996. This expectation is based on the fact that rainfall is below average during ENSO and La Niña years, and thus should be somewhat wetter during normal years. Tropical cyclone activity is expected to return to normal during 1996. Threats will be greatest at their normal time during October-December 1996. A developing tropical depression, storm, or minimal typhoon could affect the state during April and/or May as well.

POHNPEI STATE: Summertime rainfall on Pohnpei island was slightly below the long-term average, but October rainfall was about 25% above the long-term average for that month. Rainfall on Pohnpei and most surrounding islands is expected remain near-normal throughout the coming year. Pingalap may have slightly below normal rainfall until about January 1996, since it usually receives more rain during ENSO events (See Kosrae discussion). Tropical cyclone activity in 1995 was greatly reduced in Pohnpei and eastern Micronesia because of the westward displacement due to La Niña conditions. With the expected return to normal climate conditions in coming months, tropical cyclone activity in 1996 should return to normal. A developing tropical depression could affect the region during the spring, but greatest threats for Pohnpei State are expected at their normal time, during October-December 1996.

KOSRAE STATE: Rainfall for Kosrae has been about 20% less than the long-term average during the summer and fall of 1995, and is expected to remain about 20% below average through January 1996. This is because the long-term average for these months includes the wetter-than-normal conditions of ENSO years. Since ENSO conditions have ended and La Niña conditions have developed, rainfall is expected to be somewhat below average until about March, when rainfall is expected to return to normal. Tropical cyclone activity in 1995 was greatly reduced near Kosrae and other eastern areas, because of the westward displacement due to La Niña conditions. With the expected return to normal climate conditions in coming months, tropical cyclone activity in 1996 should return to normal, with greatest threats to Kosrae being in October-December 1996.
- source: UOG-WERI

PALAU: Rainfall from July to mid-August was below the long-term average, partially due to the La Niña conditions that kept monsoon activity to the west in the South China Sea. However, September rainfall was 30% above the long-term average and October rainfall was 10% above average. This above-average rainfall resulted from two short surges of monsoon activity during September, and stronger monsoon conditions during October due to Tropical Storm Zack and Typhoon Angela. Rainfall is expected to decrease in November and be near-average for the rest of 1995, through April 1996. After April, somewhat wetter-than average conditions are expected for the rest of the year. Although tropical storm and typhoon activity has been at a near-record low for 1995, activity in the rest of the current season could still threaten Palau through mid-December. With the expected return to normal climate conditions in 1996, tropical storm and typhoon activity should also return to normal. Greatest threats to the region near Palau would be expected for October through mid-December of 1996, along with some chance for tropical storm activity in the spring.
- source: UOG-WERI

MARSHALL ISLANDS: The 3-month average rainfall at Majuro for August-October was just about normal at 12.59 inches. However, the month of October was about 25% drier than normal. Due in part to La Niña conditions, the eastward extension of the monsoon trough into the Marshall Islands region did not occur as normally expected for October and November. As a result, rainfall may be about 20% below average for the remainder of 1995, and somewhat below average until about March 1996, when climate conditions and rainfall patterns are expected to return to near-normal. Also due to the lack of monsoon conditions, tropical storm and typhoon activity in 1995 will be below-normal for the RMI, especially in eastern areas. With the expected return to near-normal climate conditions in 1996, tropical storm and typhoon activity should also return to normal. Greatest threats to the RMI region will be from October through December of 1996.
- source: UOG-WERI


The information contained in the LOCAL VARIABILITY SUMMARIES section and elsewhere in this issue of the Pacific ENSO Update has been drawn from many sources. Further information may be obtained by contacting your local National Weather Service Office, or the individuals and institutions listed below:

NOAA National Weather Service - National Centers for Environmental Prediction - CLIMATE PREDICTION CENTER (CPC):
World Weather Building, Washington D.C. 20233.
Contact CPC at 301-763-8167 for more information on the ENSO Advisory, the Long-Lead Outlook for the Hawaiian Islands, and other NOAA publications discussed in this bulletin.

NOAA National Weather Service - Pacific Region WEATHER SERVICE FORECAST OFFICE (WSFO)
University of Hawaii - Manoa Campus
HIG #225, 2525 Correa Road, Honolulu, Hawaii 96822
Contact the WSFO at 808-973-5270 for more information on NWS-PR sources of climate information.

Lower campus, University of Guam
UOG Station, Mangilao, Guam 96923
Contact C. Guard or M. Lander at (671)735-2685 for more info on tropical cyclones and climate in the Pacific Islands.

University of Hawaii (UH) School of Ocean and Earth Science and Technology (SOEST) DEPARTMENT OF METEOROLOGY:
HIG #331, 2525 Correa Road, Honolulu, Hawaii 96822
Contact Dr. T. Schroeder, Chairman, at 808-956-7476 for more information on hurricanes and climate in Hawaii.

HIG #331, 2525 Correa Road, Honolulu, Hawaii 96822
Contact C. Yu at 808-956-7110 for more information on ENSO-related climate data for the Pacific Islands.
Contact A. Hilton at 808-956-2324 for more information on Pacific ENSO Update and applications.


NEW: Experimental Forecasts for Pacific Island Rainfall

Appendix I: ENSO Advisory of October 18, 1995

For further information, please contact:

Alan C. Hilton, LT/NOAA
Editor, Pacific ENSO Update,
Pacific ENSO Applications Center
c/o Dept. of Meteorology, HIG Room 331
University of Hawaii - Manoa Campus
2525 Correa Road - Honolulu, HI 96822
Tel: 808-956-2324 Fax: 808-956-2877

Publication of the Pacific ENSO Update is funded in part
by Grant Number NA46GP0410 from the National Oceanic
and Atmospheric Administration (NOAA) Office of Global
Programs. The views expressed herein are those of the
author(s) and do not necessarily reflect the views of NOAA
or any of its sub-agencies.