Home   Research   Supplementary   Personal   CV

Reconstructing Large Scale Atmospheric Circulations

My main research interests lie in the branch of tropical meteorology and the atmospheric general circulation. Mainly, I use different mathematical and statistical techniques to break down different atmospheric variables, and study them in a more simplified way. By doing this, I hope I can shed light on some fundamental aspects of the atmospheric system that might have been previously missed, or that models fail to represent. Most of my research has focused on the Madden Julian Oscillation, but have recently extended it to other modes of variability, such as the El Niño Southern Oscillation, the climatological mean stationary waves and convectively coupled equatorial waves.

Of the several techniques I've used to further my research goals lie the use of a general solution to the calculation of velocity potential and streamfunction that works on limited domains on a sphere (Adames et al. 2013) and a deconstruction of the vertical strucutres of different modes of variability through the use of matrix methods such as EOF analysis and MCA (Adames and Wallace, 2013).

The Madden-Julian Oscillation

Of particular interest are the different modes of tropical variability, with a focus on the Madden Julian Oscillation (MJO) . The MJO is an atmospheric event that features a large-scale center of deep convection that propagates eastward at a speed of around 5 m/s flanked by zones of convection and precipitation. A sample animation of how the MJO looks like in the pressure (shading) and zonal wind (contours) is shown to the right.

Four-panel schematic depicting the mid-latitude response to convection over the Maritime Continent.
 


Upper level signature of the MJO. Contours denote zonal winds (black positive, grey negative) and shading denotes upper level heights (pressure), with blue denoting lows and orange denoting highs.


Schematic of a typical extratropical response to MJO heating centered near Indonesia (phase 4). From Adames et al. (2013).

Its signal results from the interaction of various processes on different scales, from individual mesoscale convective structures to large-scale dynamics at the scale of ocean basins. The mechanisms through which those various scales interact are not yet fully understood.

The MJO modulates tropical precipitation and is directly associated with flooding events over the Indo-Pacific warm pool as well as modulation of tropical cyclone activity throughout the tropics. Through its convective and dynamic signature, the MJO affects the Earth's angular momentum, the length of day, as well as extratropical activity. It has been documented that the MJO is indirectly associated with extreme precipitation events over the west coast of North America.

In spite this great impact, the mechanism in which the MJO initiates, propagates and creates such a broad range of impacts is not well understood. Models fail to simulate it correctly, and improvement in the forecasting of the MJO has been slow. It is clear that our understanding of the strucutre and propagation of the MJO is still quite limited. This has motivated me to reanalyse the structure of the MJO.

 

One of the projects I am currently working on relates to the use of Empyrical Orthogonal Function analysis and Maximun covariance analysis through the use of Singular Value Decompositionto to separate the different structures of the MJO. This method could potentially yield new insight on how the extratropical wave pattern is generated by the MJO, for example. For more information read (Adames and Wallace, 2013).

We also study the relative contribution of the irrotational, non-divergent and environmental (deformation) components of the flow to the dynamcics of the MJO and other coupled equatorial waves by following a wind partitioning technique described by Bishop (1996) and adapted by Patoux et. al. (2005). We adapt this technique to work with global scale, tropical flow. This method adds addtional details to large-scale tropical dynamics that might not be visible other methods such as potential vorticity.