Methodology
The FSU models, coupled and nested regional models, will be used for the study of the North American Monsoon. The coupled atmosphere-ocean modeling system follows two recent papers, Krishnamurti, et.al. (2000a) and LaRow and Krishnamurti (1998). The nested regional model is described in Cocke and LaRow (2000). Figure 1 provides an outline of the global modeling system.

A unique part of the FSU modeling system is our data assimilation system. The first phase of data assimilation invokes 10 years of daily surface winds based on ECMWF analysis. That plus the Reynold’s 10 day averaged sea surface temperatures are used in an ocean spin up. The coupled data assimilation is based on the physical initialization described in Krishnamurti et. al. (2000a). Here a number of atmospheric/ocean variables such as the rotational part of the wind, the temperature field, the surface pressure field, the ‘observed’ rain rates and the sea surface temperatures are subjected to Newtonian relaxation. No flux corrections are used in the coupled seasonal forecasts.

Details of proposed work
Figure 2 illustrates a modeling stream that we have used recently for our studies of the El Niño of 1997-98. This includes an ocean spin-up phase, a coupled assimilation phase and a free forecast phase. In addition to what is described here, we propose to add an ensemble forecast component with the regional high resolution spectral model.

To better understand the interannual variability of the monsoon, we propose to map the characteristic differential heating during the summers of 1997 and 1998, periods of active and post-El Niño Southern Oscillation (ENSO) events. In addition, we propose to examine the diurnal change inherent to the North American monsoon, including its monthly variation, amplitude and phase, and to demonstrate the role of the Sierra Madre Occidental on the precipitation regime of the monsoon. Modeling experiments of the monsoon will employ the Florida State University Global Spectral Model (FSUGSM), which will allow for accurate modeling of the differential heating components as well as of the motion, temperature, and moisture fields. Differential heating between land and ocean is the basic driving mechanism of monsoon systems. Through analysis of the response of the differential heating components to the ENSO phenomenon, we hope to better understand the forces governing interannual variability of the North American monsoon. Our preliminary seasonal model forecasts of the monsoon show that northwestern Mexico and the southwestern United States experienced a decrease in monsoonal precipitation during 1997, the summer of the most recent ENSO event, in comparison to precipitation during 1998, a post-ENSO period. We would like to extend these investigations to encompass observational analyses and detailed modeling of the differential heating, moisture, temperature and wind fields. The primary objective of our research is to investigate the temporal and spatial variability of the North American monsoon via coupled atmosphere-ocean modeling and observational analyses

Our research has an observational and a modeling component. We shall be making 6 seasonal forecasts with our global and regional high resolution model for the summer seasons of 1986, 1988 and 1990 (the wet years) and 1973, 1978 and 1993 (the dry years). Each of these experiments takes an eleven years of daily data assimilation (i.e. 10 years of ocean spin up and one year of coupled assimilation). For each of these seasonal forecasts, we furthermore propose to carry out 8 ensemble forecasts for each experiment. These would be based on adjacent start dates for the assimilated data sets, i.e. May 27, 28, 29, 30, 31 and June 1, 2 and 3 for each of the respective six years. Monitoring of the following components of the North American monsoon would be carried out from both the coupled model results and from the reanalysis data sets.

I. The behavior of the subtropical ridge for the 6 experiments

II. The relationship of the North American monsoon to the Pacific Decadal oscillation for the 6 experiments.

III. Evaluate the phase and amplitude of the diurnal component

IV. Mapping of the differential heating of the North American monsoon

V. Interannual differences in the patterns of North American rainfall during the wet and dry years.

Results and Accomplishments
For the NWP modeling issues of the North American monsoon, Ms. Connie Klimczak, a graduate student, has addressed the diurnal change and the distribution of heat sources and sinks. This work is essentially completed and Ms. Klimczak expects to receive her Masters degree in this semester. She has shown that the North American monsoon does have definable heat sources and sinks, but the scale of those are much smaller than that of the Asian monsoon. She furthermore notes a large diurnal amplitude for the heating and the divergent winds. Unlike the Asian monsoon, the divergent kinetic energy is not passed on to the rotational kinetic energy. Thus the North American monsoon appears to be largely a local divergent monsoon driven by local differential heating. Fig. 3 illustrates our preliminary results on the energetics of NWP based data sets (ensemble of one day forecasts). The large (rotational) monsoon gyres similar to those of the Asian monsoon do not evolve in the same (large scale) manner. Although gyres similar to those of the Asian monsoon are present both to the east and west of the Sierra Madre Occidentale, such as the Bermuda high and the east Pacific subtropical high, they are not monsoon circulations in the same energetics sense.

The seasonal climate modeling of the North American is an ongoing project here under this grant. It entails the following components:

VI. An ocean spin with the global ocean model

VII. Coupled assimilation covering 10 years of daily input for atmosphere and ocean and High resolution regional spectral climate forecasts.

VIII. Among these three items, i) and ii) are almost completed during the first year of this project. We have now a data base to carry our regional spectral seasonal forecasts.

Future Work
During year two we will be carrying out ensemble forecasts with this system. We are transi-tioning towards the use of the FSU multimodel superensemble for seasonal to multiseasonal forecasts (see Krishnamurti, et.al., 2000b). We will also have those forecasts available for comparison with the forecasts based on an ensemble of forecasts from a single model. Our goals in these experiments will be to provide a better understanding of the North American monsoon system during an entire season over several years.

T.N. Krishnamurti
tnk@met.fsu.edu
phone: (850) 644-2210
fax: (850) 644-9642