Conducting short duration field programs to evaluate sounding site representativness and potential climate monitoring biases—Examining the Low-Level Jet Over the Venezuelan Ilanos During the 2005 Dry Season

It is often difficult to obtain good agreement between upper air observations in the tropics, made by radiosonde or pilot balloons, and output from a data assimilation system, such as the NCEP reanalysis. Sometimes the observations are affected by local conditions or topography which the large-scale analyses do not resolve. However, to determine a “true” 24-h mean wind at a location for comparison of the mean values produced by a data assimilation system requires observations throughout the diurnal cycle - which is very rarely carried out - even for short periods. In Latin American most radiosonde sites make only one observation daily. As part of the PACS-SONET effort to monitor winds over Latin America with inexpensive once-daily pilot balloon observations, we have been evaluating the uncertainty in assessing daily mean winds from these once-daily observations. Observations from some sites have shown interesting features, such as localized low-level jets, but such features cannot be easily compared with monthly mean analyses produced by the assimilation systems.

One way to estimate the bias of a particular sounding time in generating a mean daily wind is by simply making more observations per day. While this is not feasible for long periods, pilot balloon observations, because of their low individual cost, can be made frequently for short periods. With these high frequency measurements the bias in inferring mean winds from one observation per day can be estimated. This procedure will be shown to be effective in resolving the diurnal variability of a strong dry season low-level jet (as judged from morning only soundings) over the Venezuelan llanos. We requested twice-daily observations to be made for one month, then made special frequent soundings during a shorter period to resolve the phase of the diurnal cycle of the winds. The details of this activity, carried out in February and March 2003, will be presented. In addition to resolving better the diurnal wind variability as a function of height, the Venezuelan activity consists of observations being made at surrounding sites, allowing us to estimate the horizontal scale of the jet. The steadiness of the dry season jet allows us to extend our conclusions to much of the dry season,; the validity of our conclusions would have to be verified at other times of the year, especially the wet season. Finally, the feasibility of doing this at every sounding site in tropical regions will be discussed.