Daniel J.
Vimont
Home
Research
Class
Miscellaneous
Links
Ongoing Research Projects
Interactions between the midlatitudes and tropics: the Seasonal Footprinting Mechanism
Analysis of several long
simulations using the CSIRO coupled general circulation models has
identified a mechanism by which the mid-latitudes can affect the
tropics, through coupling with the oceanic mixed layer in the
subtropics and tropics. During the winter season (when the North
Pacific atmosphere is most energetic), mid-latitude atmospheric
variability can affect the strength of the subtropical trades, which
can impart an SST "footprint" onto the ocean surface through variations
in the net surface heat flux. These SST anomalies persist beyond
the life time of the mid-latitude variability, and can influence the
tropical and subtropical atmospheric circulation during the ensuing
seasons. The residual atmospheric response to the SST footprint
includes zonal wind stress anomalies along the equator, which can give
rise to ENSO variability in the CSIRO models.
Further research has identified this "seasonal footprinting mechanism" in the observed record, and has illustrated the potential for positive coupled feedbacks between the off-equatorial SST anomalies and the residual atmospheric circulation that they force. Results from the observed record indicate that the SFM may be an important contributor to ENSO predictability.
We are currently working with Michael Alexander (at the NOAA Climate Diagnostics Center in Boulder, CO) to explore feedbacks between the SST footprint and atmospheric variability that it generates, using a suite of coupled general circulation models. The experiments are designed to quantify the role of realistic ocean physics (e.g., mixed layer processes, ocean dynamics) on the evolution of ENSO variability. This research is being funded by a grant from the NOAA Climate and Global Change Program, through CLIVAR.
Further research has identified this "seasonal footprinting mechanism" in the observed record, and has illustrated the potential for positive coupled feedbacks between the off-equatorial SST anomalies and the residual atmospheric circulation that they force. Results from the observed record indicate that the SFM may be an important contributor to ENSO predictability.
We are currently working with Michael Alexander (at the NOAA Climate Diagnostics Center in Boulder, CO) to explore feedbacks between the SST footprint and atmospheric variability that it generates, using a suite of coupled general circulation models. The experiments are designed to quantify the role of realistic ocean physics (e.g., mixed layer processes, ocean dynamics) on the evolution of ENSO variability. This research is being funded by a grant from the NOAA Climate and Global Change Program, through CLIVAR.
Pacific ENSO-like decadal variability
A
defining characteristic of tropical Pacific decadal variability is its
resemblance to interannual ENSO variability (see panels (a) and (b) of
the figure to the right). Two noticable differences are the
meridionally broadened SST anomalies in the tropical Pacific for
decadal time scales, and the increased magnitude of mid-latitude SST
anomalies (relative to tropical SST anomalies) on decadal time scales,
compared to interannual time scales. To explore possible mechanisms of tropical Pacific "ENSO-like" decadal variability, we attempt to reconstruct the decadal ENSO-like SST pattern using a linear combination of interannual spatial patterns. So, we reconstruct the decadal SST using only the first four EOF's of interannually filtered SST over the Pacific. This new data set has spatial information from interannual time scales only (the interannual EOFs). We then apply EOF/PC analysis to the reconstructed decadal SST data, and identify a spatial pattern of variability (panel (c)) that bears a strong resemblance to the decadal ENSO-like pattern in panel (a). Results indicate that decadal processes may not be necessary to explain the ENSO-like pattern of decadal variability.
What mechanisms are responsible for generating the decadal pattern of variability? Examination of the interannual EOFs that are used to reconstruct the decadal pattern in panel (c) (above) indicate that only three interannual EOFs are necessary to represent the pattern of ENSO-like variability with fidelity (only these three are used in calculating the map in panel (c)). These EOFs represent ENSO precursors (via the seasonal footprinting mechanism, described above), the peak of an ENSO event (panel (a)), and ENSO "leftovers" (generated by the atmospheric bridge, and by equatorial waves propagating along the oceanic thermocline). The results suggest that the ENSO-like pattern results from averaging over the "debris" of interannual ENSO variability. This is important, because any process that can affect the decadal variability of ENSO (e.g., coupled processes involving mid-latitude atmospheric variability) should generate this ENSO-like response, when averaged over a long enough time period.
Downscaling Indonesian Precipitation
This image of eastern
Java was taken by the
Moderate Resolution Imaging Spectroradiometer (MODIS) satellite on
October 19, 2002. Red dots indicate fires that were detected by
the MODIS satellite (smoke plumes are evident in the photo). This
picture was taken following several months of unusually dry conditions
associated with the warm ENSO event of 2002-2003. Click on the
photo for higher resolution options, where you can see internal waves
in ocean to the southeast of Bali, in the lower right-hand corner of
the picture.
As the picture above illustrates,
precipitation over Indonesia is strongly affected by ENSO variability
in the tropical Pacific. During ENSO warm events Indonesia
experiences dry conditions that delay rice plantings. This has a
large impact on Indonesia's economy.
This work began with Dan Schrag (at Harvard University) while I was working under a short post-doctoral appointment at Columbia University's Earth Institute. During that appointment, we began the downscaling investigations, and built the framework for the coupled model simulations.
We are currently collaborating with economists Rosamond Naylor and Walter Falcon at Stanford, and David Battisti at the University of Washington to identify how ENSO can impact precipitation on a regional scale in Indonesia under present and future climate states. These relationships will be used to identify food security in Indonesia, and potential steps that can be taken to minimize risk to Indonesia's economy.
In Indonesia, seasonal precipitation is controlled by the annual cycle of the monsoon, and variations associated with ENSO. General circulation models (GCMs) are capable of reproducing many features of the large scale circulation (such as the Southern Oscillation, or the monsoon shear line), but have trouble simulating the hydrological cycle. On regional scales, precipitation is strongly influenced by interactions between the large scale circulation and small scale topography (which GCMs cannot resolve). Empirical downscaling models (EDMs) are generated that relate the large scale circulation to regional hydrology using relationships in the observed record. We apply these EDMs to GCM simulations of present and future (greenhouse gas forced) climate to produce downscaled and debiased precipitation scenarios for Indonesia.
This work began with Dan Schrag (at Harvard University) while I was working under a short post-doctoral appointment at Columbia University's Earth Institute. During that appointment, we began the downscaling investigations, and built the framework for the coupled model simulations.
We are currently collaborating with economists Rosamond Naylor and Walter Falcon at Stanford, and David Battisti at the University of Washington to identify how ENSO can impact precipitation on a regional scale in Indonesia under present and future climate states. These relationships will be used to identify food security in Indonesia, and potential steps that can be taken to minimize risk to Indonesia's economy.
In Indonesia, seasonal precipitation is controlled by the annual cycle of the monsoon, and variations associated with ENSO. General circulation models (GCMs) are capable of reproducing many features of the large scale circulation (such as the Southern Oscillation, or the monsoon shear line), but have trouble simulating the hydrological cycle. On regional scales, precipitation is strongly influenced by interactions between the large scale circulation and small scale topography (which GCMs cannot resolve). Empirical downscaling models (EDMs) are generated that relate the large scale circulation to regional hydrology using relationships in the observed record. We apply these EDMs to GCM simulations of present and future (greenhouse gas forced) climate to produce downscaled and debiased precipitation scenarios for Indonesia.
Tropical Meridional Modes
In the Atlantic, coupled climate
variability is dominated by variability with a meridional
structure. Variations in the meridional SST gradient drive
surface winds that flow towards the warmer hemisphere. This
impacts the location of the ITCZ, with large impacts on drought in the
Nordeste region of Brazil, and in the Sahel region of Africa.
In collaboration with John Chiang at the University of California, Berkeley, we have identified a similar "meridional mode" of coupled ocean-atmosphere variability in the tropical Pacific. While not identical, features of the Pacific meridional mode share some strong similarities with the Atlantic meridional mode. Most notable is the meridional SST gradient, meridional surface winds that maximize over the maximum meridional SST gradient, and an ITCZ shift towards the warmer hemisphere. It appears that both modes can be excited by atmospheric variability over their northern basins, and may play a role in the seasonal cycle. The identification of the meridional mode in both the Pacific and Atlantic suggests that it is a fundamental mode of the climate system. Ongoing research is aimed at identifying meridional mode variability in other basins, as well as defining the dynamics and thermodynamic feedbacks associated with its existence.
In collaboration with John Chiang at the University of California, Berkeley, we have identified a similar "meridional mode" of coupled ocean-atmosphere variability in the tropical Pacific. While not identical, features of the Pacific meridional mode share some strong similarities with the Atlantic meridional mode. Most notable is the meridional SST gradient, meridional surface winds that maximize over the maximum meridional SST gradient, and an ITCZ shift towards the warmer hemisphere. It appears that both modes can be excited by atmospheric variability over their northern basins, and may play a role in the seasonal cycle. The identification of the meridional mode in both the Pacific and Atlantic suggests that it is a fundamental mode of the climate system. Ongoing research is aimed at identifying meridional mode variability in other basins, as well as defining the dynamics and thermodynamic feedbacks associated with its existence.
Climate Variations and Hurricane Activity
Composite SST (shading) and vertical wind shear (contour 0.25 m s-1; solid positive, dashed negative) anomalies around the five years with the positive (left) and negative (right) values of the Atlantic Meridional Mode (AMM) during July-November. Also shown are the genesis dates of tropical storms (+’s). Storms that, at some point in their life, achieve major hurricane status (class 3 or above) are indicated by a circle around their genesis point.
Recent work with Jim Kossin at the
University of Wisconsin has identified a strong relationship between
seasonal hurricane activity and the Atlantic Meridional Mode
(AMM). In a series of three papers it is found that: (i)
previous findings of upward trends in global hurricane activity are not
supported when reanalyzed using a consistent satellite-based retrieval
of hurricane intensity – the only basin with a significant upward trend
is the Atlantic; (ii) in the Atlantic, the AMM exerts a strong
influence on a number of climatic conditions that all cooperate in
their influence on hurricane activity; and (iii) the AMM provides a
more general framework for understanding hurricane / climate
interactions through its influence on storm frequency and duration, and
hence indirectly on average intensity (this is evident in the figure
above).
The preliminary analysis summarized above provides strong motivation for future research into hurricane climate interactions. In particular, we intend to (i) initiate a comprehensive set of modeling and observational studies to identify the physical processes that link the AMM to the suite of large-scale conditions that influence hurricane activity; (ii) undertake a consistent global analysis of hurricane / climate interactions across all ocean basins; (iii) identify potential hurricane season predictability through potential predictions of the AMM (preliminary results indicate that predictions of the AMM a year in advance nearly triples the explained variance of storm frequency over conventional forecast metrics); and (iv) identify whether a new framework can be developed to investigate how hurricane activity could change with global warming.
The preliminary analysis summarized above provides strong motivation for future research into hurricane climate interactions. In particular, we intend to (i) initiate a comprehensive set of modeling and observational studies to identify the physical processes that link the AMM to the suite of large-scale conditions that influence hurricane activity; (ii) undertake a consistent global analysis of hurricane / climate interactions across all ocean basins; (iii) identify potential hurricane season predictability through potential predictions of the AMM (preliminary results indicate that predictions of the AMM a year in advance nearly triples the explained variance of storm frequency over conventional forecast metrics); and (iv) identify whether a new framework can be developed to investigate how hurricane activity could change with global warming.