You are here: Home / Research / Groups / Echanges Côte-Large / Regions of interest / South-East Pacific / Atmospheric statistical downscaling

Atmospheric statistical downscaling

by ECOLA last modified Oct 13, 2014 11:20 PM

Atmospheric statistical downscaling and impact on oceanic circulation

G. Cambon, K. Goubanova (post-doc), P. Marchesiello, B. Dewitte, S. Illig (LEGOS/IRD)

Long-term simulation of oceanic circulation in Eastern Boundary Upwelling Systems (EBUS) is a challenging issue due to the paucity of wind stress products of a sufficiently high spatial resolution. To circumvent this resolution issue, we need to use atmospheric downscaling techniques, either dynamical or statistical. While dynamical downscaling may be cumbersome due to computational costs associated with the use of high-resolution mesoscale atmospheric models, statistical downscaling offers a cost effective alternative. In the Peru–Chile coastal ocean, Goubanova et al. (2011) showed that statistical downscaling of reanalysis products provide improved coastal winds over extensive time periods. We assess the value of this atmospheric downscaling on the oceanic circulation through twin regional simulations of the Humboldt currents system (Peru and Chile coasts). The twin simulations only differ from wind stress forcing over the 1992-2000 period that encompasses the major 1997/98 El Niño/La Niña events. It is shown that the mean biases of the oceanic circulation can be drastically reduced simply substituting the mean wind field of NCEP reanalysis by a higher resolution mean product (QuikSCAT). The statistical downscaling model further improves the simulations allowing for more realistic intraseasonal and interannual coastal undercurrent variability, which is notoriously strong off Central Peru and Central Chile. Despite some limitations, our results suggest that the statistical approach may be useful to regional oceanic studies of present and future climates.


Figure: Vertical sections of meridional current (in cm.s–1) off Central Peru (alongshore-averaged between 7_S and 13_S) during the El Niño peak phases (December 97 to February 98) for the various simulations. Contour interval is 1 cm.s–1. The thick blue line superimposed on the color plot is the 2 cm.s–1 isoline of mean meridional current and indicates the core of the mean PCUC.

Cambon G., K. Goubanova, P. Marchesiello, B. Dewitte, S. Illig and V. Echevin, 2013: Assessing the impact of downscaled winds on a regional ocean model simulation of the Humboldt system, Ocean Modelling, 65, 11-24, doi:10.1016/j.ocemod.2013.01.007

Document Actions

logo cnes logo IRD Logo université de Toulouse Logo université Paul Sabatier Logo CNRS
Logo bibliothèque OBS Logo Observatoire Midi Pyrénées