A dynamical interpretation for the evolution of the equatorial components of the mountain torque

The evolution of the two equatorial components of the mountain torque (EMT) is analysed using 29 years (1979-2007) of the NCEP reanalysis. We find a very strong lagged correlation between its component along the Greenwich axis (EMT1) and its component along the 90°E axis (EMT2): a positive EMT1 is followed very significantly by a negative EMT2. This relationship is significant well above the 99% level for periodicities between 2 and 30 days. The physical meaning of this phase quadrature is that the EMT vector has a strong tendency to rotate westward at all periodicities, including the very fast frequencies. This makes difficult to explain this rotation in terms of planetary scale atmospheric dynamics only.

We therefore analyze the individual contributions of the major mountain ranges (Antarctica, the Himalayas, Greenland, the Rockies and the Andes) to the EMTs. The equatorial torques created by the Antarctica are dominant, (explaining about 50 to 80% of the total torques variance depending on the season), followed by the Himalaya's (15-20%) and the Rockies (10-15%). The Andes, Greenland and the other mountain ranges have an even smaller contribution to the EMTs. It is striking that, for each of these mountain range (except Greenland), the individual contribution to the global EMT exhibits the same behavior as the global EMT itself, e.g. a westward rotation in a broad frequency range.

Composite analyses of surface pressure maps keyed to the two equatorial components of the regional Mountain Torques are then used to explain this rotation. If for Antarctica, the equatorial torques evolution is associated to planetary scale modes of motion, their evolution for the Himalayas and the Andes is associated to much more local features. Typically, and for the Himalayas, a positive EMT1 is associated with a high pressure anomaly to the North. This anomaly thereafter moves south-eastward creating a very significant negative signal on EMT2. To a large extent, this behaviour is intimately related to the “cold surges” that modulate the weather in winter over East Asia. A quite comparable interpretation holds for the Andes and the Rockies.

We then argue that this westward motion of the equatorial mountain torque components witnesses of a dynamical influence of the mountain ranges on the atmosphere. With a theoretical model for the linear atmospheric response to a mountain forcing on the sphere, we show that an eastward flow produces a negative EMT along the equatorial axis that is at 90° to the east of the dominant longitude of the mountain range. The lee cyclone development that follows produces a positive mountain torque along the Equatorial axis corresponding to the longitude of the mountain. If we take for the mountain profiles in our theoretical model, profiles near the observed ones, the equatorial torques produced compare well those observed.