Meridional moisture transport by tropical synoptic scale disturbances over the Pacific and Atlantic basins

Synoptic scale disturbances in the tropical Pacific and Atlantic basins are generated from easterly waves and by the vortex roll-up mechanism acting on the ITCZ, which on time scales of a few days can break the ITCZ down into tropical disturbances. The disturbances generated by the vortex roll-up mechanism are weak and only a few of them will intensify into tropical storms. Since tropical disturbances are generated near continuously by either or both mechanisms, it has been suggested that they may play a role in meridional moisture transport. The goal of this study is to quantify the moisture transport in the two ocean basins and discuss differences between the two regions.

The specific humidity in the two analyses, ERA and NCEP, during May--July 2004 is evaluated by comparing it to AIRS measurements. ERA data are of T106 horizontal resolution (1.125degx1.125deg) and NCEP data are of 1degx1deg resolution. Both data sets contain four records per day. After the evaluation, the meridional moisture transport by synoptic scale disturbances in the two datasets is calculated by using data from May--Oct 2000--2004. May to October is the active season for tropical disturbances in these regions.

Both analyses show good agreement with AIRS near the surface over the oceans, and high bias (both mean and variance fields) in the ITCZ over land. Basically, they match the moisture distribution quite well, except in central America and the Andes mountains (where the mountains are very narrow and are not at all well depicted in the model). NCEP analysis has a systematic high bias in both the mean and variance of specific humidity in the ITCZ and SPCZ above 850 hPa, but the spatial distribution is in good agreement with AIRS. ERA data show even higher biases than NCEP at lower levels, but at upper levels the biases are lower, meaning that the moisture in ERA data is confined to the lower troposphere while the moisture in NCEP is more spread out over the entire atmospheric column. After comparing AIRS, NCEP, and ERA data, we think that the specific humidity field in NCEP analysis is more consistent with AIRS data, and is of better quality for this study. However, we should keep in mind that the time period for which AIRS data is available is short so that this comparison is based on a 3-month dataset.

Using the two analyses from May--Oct, 2000--2004, two approaches are used to estimate the meridional moisture transport. The first approach is to compute the stationary and transient components of the moisture transport. The stationary term is the 6-month mean and the transient component is the deviation from the mean. The second approach is to apply a 2--20 day filter on the wind and moisture fields to isolate the synoptic timescale.

Our results show that a comparable amount of moisture is transported out of the tropics by synoptic scale disturbances in the two tropical ocean basins. The tropical synoptic scale moisture transport is only significant in the eastern tropical ocean basins. This is because only very few of the tropical disturbances last long enough to propagate into the western ocean basins. The amount of moisture transported has larger interannual variability in the eastern Pacific than they do in the eastern Atlantic. This is because the synoptic scale disturbances in the eastern Atlantic mostly originate as African easterly waves, which are strongly controlled by the land-ocean contract. By contrast, in the eastern Pacific, many ITCZ breakdown events are due to the vortex roll-up mechanism which is more variable in terms of the location and the intensity of the disturbances generated from the breakdown. Both analysis data show similar results in the vertical integrated meridional moisture transport by the synoptic scale disturbances.