Observational Study of the Remote Forcing of the Pacific Subtropical Highs

1
Observational Study of the Remote Forcing of the
Pacific Subtropical Highs

• Richard Grotjahn and Sheri Immel Department of
  Land, Air, and Water Resources University of
  California, Davis, CA, 95616-8627 U.S.A.

2
Terminology and Abbreviations

• NP high North Pacific subtropical high.
• SP high South Pacific subtropical high.
• P precipitation
• SLP sea level presssure
• r rank correlation coefficient
• MA anomaly from averages for that month
• OLR outgoing longwave radiation
• EP flux Eliassen-Palm Fluxes
• b meridional derivative of Coriolis parameter
• v meridional component of velocity (lt0 is
  southward)
• MJO Madden-Julian Oscillation
• ENSO el nino Southern Oscillation

3
Introduction

• NP high and SP high have different magnitude and
  timing of seasonal change. See Fig. 1.
• The highs are forced by local and remote
  mechanisms that foster sinking above the high
• Three remote mechanisms have been proposed.
  (discussed separately below)
• This work has modest goals (currently unfunded)
• Primary goal of this study to test if proposed
  remote forcing mechanisms are likely or not
  likely to play a role in subtropical high
  strength using very simple tests.

4
Current Mechanism

• Rising motion in deep tropical convection in
  target region both Eastward and equatorward of
  the highs sets up a pressure pattern and/or
  circulation that fosters sinking on the east side
  of the high.
• That sinking enhances equatorward surface winds
  from the proposed balance b v -dw/dz
• Simplest test is to see if precipitation in
  target region is associated with subtropical high
  strength.
• Relevant references Hoskins Rodwell (1995),
  Hoskins (1996), Hoskins et al. (1999)

5
Classical Mechanism

• Rising motion in deep tropical convection
  Westward and/or equatorward of the highs leads to
  Hadley and/or Walker type circulations with
  sinking over the highs
• Simplest test is to see if precipitation in
  various target regions are associated with
  subtropical high strength.
• Relevant references extension of textbook
  descriptions of zonal average Hadley
  circulation to zonally-varying highs, e.g.
  Grotjahn (1993). See also Chen (1999)

6
Extratropical Forcing Mechanism

• Extratropical cyclones might
• suppress the areal extent of the subtropical
  highs OR
• have advection and divergent circulations that
  enhance the highs.
• Simplest tests look in storm track regions for
  shifts of SLP and P fields.
• Relevant references none? But, many references
  find connections between midlatitude eddy forcing
  and stationary wave properties particularly at
  upper levels. Zonal mean cells e.g. Pfeffer
  (1981). Lau Holopainen (1984) include surface
  features for the winter NP high

7
Analysis Procedures

• Preliminary study to identify coincident
  behavior.
• Monthly NCEP/NCAR Reanalysis data (1979-97).
• Seasonal groupings, local summer emphasized.
• Total and monthly anomaly (MA) fields. (MA
  defined as deviations from the average
  constructed from all occurrences of that month).
• Monthly data cannot distinguish cause from
  effect.
• Tools shown here simple composites and 1-point
  rank correlations. Significance tests are
  bootstrap resampling and t- and D-statistics.

8
Composite Maps (figs. 2, 3 6)

• Separate combinations of months with strongest
  highs and months with weakest. Composites of
  strong highs shown. Difference maps strong highs
  minus weak highs also available.
• Separate composites for SP and NP highs
• MA data shown use extended season months MJJAS
  for NP high ONDJF for SP high. Total fields
  available.
• Boostrap resampling method used to test
  significance.

9
1-Point Correlations (figs. 4 5)

• Rank correlations of 1 SLP factor versus 2-D map
  of P.
• Two significance tests used. Intersection of the
  areas by both tests are shaded. These areas tend
  to correspond closely with the 0.3 correlation
  coefficient.
• Factors tested include
• SLP at a specific grid point
• Longitude of SLP maximum in the high
• Peak value of SLP maximum
• Summertime SP and NP highs considered separately.
• Total field data tends to show larger areas and
  more obvious dipolar pattern indicative of
  shifting lines of precipitation. MA and total
  results most similar along ICZ.

10
Results (page 1 of 2)

• Main Caveats
• Cause and effect yet not examined
• Significance only approximately assessed on
  composite maps
• Significant correlations and composites of P are
  found at remote locations
• Tropical western Pacific and Indonesia
• ICZ immediately equatorward of the high
• Subtropical latitudes W of each high (e.g. SPCZ)
• Middle latitudes (storm track poleward and W of
  each high)
• Tendency for each remote factor to mainly affect
  its side of the high seen in correlations for
  both total and MA data.
• Results for MA and total data similar except
  fewer shifts (fewer dipole patterns) in P
  fields with MA data.

11
Results (page 2 of 2)

• P pattern is not opposite for weak highs in
  some regions making correlation and composite
  data differ. E.g. more rain occurs SE of NP high
  for both weak and strong highs.
• Total and MA data in composites and correlations
  show SP high stronger with P deficit over
  Kiribati region and excess over Papua N.G.
  Suggests possible intraseasonal link to MJO
  and/or interannual link to ENSO.
• Evidence supporting the Current view is not
  strong it is
• weak in NH and
• unclear in SH
• Further works with daily data, improved
  diagnostic analyses and model simulations are
  needed to establish cause and effect and
  illuminate the dominant mechanisms.

12
References cited

• Chen, P., 1999 On the origin and seasonality of
  the subtropical anticyclones and cyclones.
  Submitted to J. Atmos. Sci. (Presented orally at
  last AOFD meeting in NY, 6/99)
• Grotjahn, R., 1993 Global Atmospheric
  Circulations Observations and Theories, Oxford
  Univ. Press, 390 pp.
• Hoskins, B., 1996 On the existence and strength
  of the summer subtropical anticyclones. Bul.
  Amer. Met. Soc., 77, 1287-1292.
• _____, and M. Rodwell, 1995 A model of the Asian
  summer monsoon. Part I The global scale. J.
  Atmos. Sci., 52, 1329-1340.
• _____, B., R. Neale, M. Rodwell, G.-Y. Yang,
  1999 Aspects of the large-scale tropical
  atmospheric circulation. Tellus, 51A-B, 33-44
• Lau, N.-C., and E. Holopainen, 1984 Transient
  eddy forcing of the time-mean flow as identified
  by geopotential tendencies. J. Atmos. Sci., 41,
  313-328.
• Pfeffer, R., 1981Wave-mean flow interactions in
  the atmosphere. J. Atmos. Sci., 38, 1340-1359.

13
Future Work

• Data Work with daily data and other fields,
  especially the divergent circulation and OLR.
  Supplement observations with numerical model data
  as needed.
• Observations Refine and greatly expand the
  analytical and statistical techniques. Examples
  include lag/lead correlations with significance
  tests, lag/lead composites with bootstrap
  testing, observational evaluation of terms in key
  equations (e.g. 3-D EP fluxes, as well as simple
  composites of terms), trajectory analysis.
• Modeling Design numerical model experiments such
  as sensitivity tests for remote and local
  forcing quantities, baseline studies to better
  understand observed significant events. Consider
  applications of local interest such as air
  quality. Proposed model MM5.

14
Signup/Comments

• Would you like more information? Please visit our
  website http//www-atm.ucdavis.edu/grotjahn/Subh
  i/
• A handout is available. If Im out of handouts,
  please sign up below.
• Name Email Postal Address

15
Figure 1 NP and SP Climatology

• SLP maximum value during each calendar month in
  the 19 year data record. Median value is blue
  line, values between the red and blue are in the
  3rd quartile, those below the yellow line are in
  the 1st quartile.
• Beneath are similar plots for latitude and
  longitude location using grid point numbers. The
  interval is 2.5 degrees in each direction and the
  values increase towards the E from the Greenwich
  meridian.

16
Figure 2 NP high Composites

• 6 strongest NP highs in May-Sept. periods of the
  19 year record. 6 weakest in those periods as
  well. MA data shown.
• SLP significant anomaly regions differ between
  strong and weak. They are not opposite. Strong
  highs may have some association with stronger SP
  high and SH extratropical lows (Fig. 6)
• P significant anomaly regions Central Pacific
  ICZ is closer and stronger for strong high, E.
  Indonesian P stronger for strong high, roughly
  opposite patterns for weak highs. E Pacific
  shifted S for weak highs, but even further S for
  strong highs. Central American P positive anomaly
  for both composites. Midlatitude P over Aleutians
  may be greater for stronger highs.

17
Figure 3 SP high Composites

• 6 strongest SP highs in Oct.-Feb. periods of the
  19 year record. 6 weakest in those periods as
  well. Total and MA data shown.
• Significant SLP regions are opposite between
  strong and weak high composites over Amazonia and
  the far western equatorial Pacific.
• Significant P regions entire Pacific ICZ is
  stronger further away for strong high. Total
  data show ICZ shift away for strong, towards for
  weak SP high. MA data have significant regions
  only where P is greater e.g. SPCZ stronger for
  strong highs.
• P in the Kiribati region (15 degrees N of Fiji)
  is opposite for weak and strong highs less P
  there (with ICZ SPCZ split) for strong SP high,
  more P there for weak high.
• P along extratropical storm track stronger for
  strong highs. There may be greater spread for
  weak highs.
• Presumably MJO and ENSO can enhance or reduce the
  convection in this region and thus could be
  linked to SP high strength.

18
Figure 4 NP 1-pt Correlations

• Correlations (contoured) between 2-D field of P
  and SLP at a single grid point (the key point).
  Each key point is indicated by a circled
  asterisk. Shaded areas indicate significance
  gt95. Brown colors mean less P is associated with
  higher SLP at the key point. Blue means greater P
  occurs for higher SLP at the key point. H marks
  the NP high JJA average location.
• Caution H marks high center only in total data
  not MA data.
• Higher r values tend to occur at remote spots on
  the same side of the high as the key point.
• Little of the P near Central America is
  significant, even for key points on the East side
  of the high, contrary to the current view. That
  which is is less not more.
• Total data have much larger significant areas.

19
Figure 5 SP 1-pt Correlations

• Similar to fig. 4 except for SP high in DJF
• Total data have much larger significant areas
  than MA data. r values from MA and total data
  have opposite sign over E Amazonia for NE and E
  key points. Another difference is that total data
  have more dipolar patterns implying seasonal
  shifts of the ICZ. Otherwise the r values from
  the two datasets are quite similar.
• Total and MA data both show W, SW, and S key
  points have dipolar correlation pattern implying
  poleward shift of extratropical storm track P for
  higher SLP and vice versa.
• Key points to the E, NE, N, and NW of the SP high
  center correlate with less P in the Kiribati
  region and vice versa, consistent with the
  composites results.

20
Figure 6 wider area composites

• Wider views of the composites shown in figs 3 and
  2 except that MA data are used.
• Strong summer NP highs seem to have some
  association with SH SLP.
• Strong summer SP highs do not seem to have much
  connection to NH fields of SLP and P.

21
Figure 7 OLR vs P comparison

• Comparison of composite differences (Strong highs
  minus weak highs) for two fields OLR (upper
  plot) and P (lower plot).
• Note Record length and months used differ
  between the two plots as do the individual highs
  chosen.
• This plot did not reproduce well in the preprint
  volume.