The Climate System

EESC 2100 Spring 2007

Lectures - Monday and Wednesday, 11:00 AM - 12:15 PM
Lab - Tuesday, 4:10 PM -7 PM

Case Studies: ENSO Impacts

This material builds from what we learned on Monday's lecture. Specifically we use the knowledge of how and why the tropical Pacific ocean-atmosphere system changes every 3-7 years to understand how these changes impact the climate of the regional and, in fact, the climate of regions very far away from the tropical Pacific.

From this session you should have an understanding of why El Niño and La Niña cause the climatic changes that they do, both from an intuitive sense based on your present knowledge of basic atmosphere and ocean dynamics, and a specific understanding of the natural variability of this system, its role in affecting the climate of near and far-field regions, and its related societal impacts. This recitation underscores the issue that regardless of your academic focus, year-to-year changes in the ENSO state produce very real climatic, social, and economic perturbations at a global scale.

Here are some of the known global climate impacts of El Niño (warm event) and La Niña (cold event).

Here we consider the effects of ENSO variability on the climate of SE Australia as a case study for ENSO impacts elsewhere. Then, we turn the discussion over to you and ask you to investigate what these ENSO climate linkages really mean, how reliable they are, and how they might be used in practical management decisions on the scale of an individual farmer or governmental agencies charged with allocating irrigation water.

I. Case Study of ENSO Impacts:Australia

• High potential evaporation (E) rates over entire continent (Fig 1). Semiarid setting makes it very sensitive to ENSO-related changes in regional rainfall.
• Model calculations of wheat yield in Australia, which are quite sensitive to rainfall amounts and timing during the growing season, agree quite well for a period of 18 years [1975-93] (Fig 2). This is true for each of the five states in which wheat is grown in Australia.
• Wheat is grown in a relatively narrow band of Australia where rainfall, temperature and soil conditions are optimal (Fig 3). Average (1901-99) crop yields in each shire (county) range over about a factor of eight, from 0.5 to 4.0 T/ha, with a national average of about 1.7 T/ha.
• National annual yields per unit area (simulated by model calculations) vary from less 1 T/ha to more than 2.1 T/ha during the twentieth century (Fig 4), largely in response fluctuations in rainfall amounts and timing relative to cropping seasons. A significant fraction of this interannual variability in rainfall patterns is correlated with variations in ENSO parameters.
• The percentage of variation in simulated annual wheat yields tends to be relatively small in Western Australia and relatively large in SE Australia where correlations of rainfall amounts with ENSO tend to be the strongest (Fig 5).

II. Response of River Discharge in SE Australia to ENSO

• Location of discharge gauge for combined flow of Murray and Darling Rivers (Fig 6)
• Largest discharge (Q) and most economically important river system in Australia
• All Q values discussed here have been corrected from measured values to include estimates of water loss due to
  1. evaporation from reservoirs,
  2. net diversions from the channel for irrigation or domestic supplies, and
  3. water additions by imports from adjacent basins during droughts (Snowy Mountain Scheme water diversions).

  After these corrections have been applied, the resultant Q is termed "natural Q", i.e. what we think it would have been without human influence.

• Natural annual Q in the Murray at the downstream end averages about double the current gauged (Actual) annual Q values
• Interannual variation of natural Q (factor of 15) in Murray very large (Fig 7), which presents major difficulties in providing sufficient water for irrigation and urban areas, as well as flood protection
• Hudson River in NY state has similar mean annual Q as River Murray, from a drainage basin a factor of 30 smaller (it rains more in the NE USA than in SE Australia), but a much smaller range of interannual variation (Fig 8)
• ENSO is a major cause of the large interannual changes in river discharge in SE Australia
• Both Darling and Murray rivers tend to have low Q during El Nino years, although, precipitation derived from quite different water vapor sources (Fig 9)
• About two thirds of El Nino years during last century had annual Q values less than the median.
• Murray receives winter P max from temperate zone storms, while Darling receives summer Australian monsoon P (low latitude vapor sources)
• Scatter plot of Natural Q for the Darling and Murray river vs SST for the Wright Index region shows strong influence of ENSO with drought mostly occurring during El Nino years (Fig 10)
• Correlation of Darling and Murray Q with SST (Wright) is not as strong as between SOI and SST. Thus ENSO is not the only factor influencing interannual differences in river discharge in SE Australia, although it is clearly the most important single large-scale meteorological process that has been identified up to now.

III. Questions for you to consider:

1. If the weather forecast says there is a 90%, 50% or 20% chance of rain, what does this mean? If it doesn't rain was the forecast incorrect? In yesterday's lab you computed the correlation coefficient (r) between the SOI and SST indices of ENSO. If you were told that the square of the correlation coefficient (r*r = r²) is a measure of the degree of shared variability (relatedness) between two potentially correlated variables, what value of r would you want to have to say a region had a 90%, 50%, or 20% chance of being linked to ENSO? (hint: 90% = 90/100 = 0.90).
2. What happened in India during last year's El Niño? Drought is expected in an El Niño year (Fig 11) which 1997 certainly was. And ENSO is certainly correlated with food production (Fig 12). But last year was quite normal (Fig 13) -- though not everywhere of course (Fig 14). What are some reasons why the crop yield was not affected by the strong El Niño of 1997? Consider the same question for Peru - Since there is a strong ENSO connection in Peru, what possible reasons are there for the low correlation with NINO3 of maize in Peru?
3. The maize crop in Zimbabwe was planted in October 1997. Suppose you were advising the government at that time? What would you tell them to expect? How confident are you? How would you convince them to pay attention to you? (CF Fig 15 and Fig 16). Using the r-value in Figure 5, what is the degree of shared variability (relatedness) between Zimbabwe crop yield and ENSO? (Hint: see question 1). How could you use this statistic to convince the authorities?
4. Some places have taken action in the past to mitigate ENSO impacts. In the Nordeste region of Brazil, they adjust planting dates (Fig 17). In Peru coastal farmers plant rice instead of cotton (Fig 18). Which other regions might adopt such strategies? How would this affect the future correlations between ENSO and crop yields for these places?
5. Arizona (Fig 19), one of the most arid states in the U.S.A., has two large urban centers (Phoenix and Tucson) plus major irrigation zones. Most of the municipal water supply for Phoenix is derived from climate district #4 in the central highlands. Precipitation in this district is significantly different during El Nino years, compared with La Nina years (Fig 20). Which months show the largest differences in precipitation amount between El Nino and La Nina years? What changes in management practices for the water storage reservoirs in the central highlands would be reasonable during El Nino years?

Note: ENSO conditions as of October 1998 are shown in Figure 21 which shows the global SST anomalies as of the end of that month. How does this SST pattern compare with the 1997 El Niño?

Text by Mark Cane, Peter deMenocal, and Jim Simpson.

home

syllabus

target skills

viewer tutorial

excel intro

ees

Columbia University

Dept. Earth & Env. Sci.

Barnard Dept. Env. Sci.

LDEO

Earth Institute