This study describes a satellite-based retrieval algorithm designed to obtain the atmospheric water budget over the open ocean. The algorithm requires optical visible and infrared measurements from a geosynchronous imager and passive microwave measurements from a polar orbiting satellite. To test the methodology, a combination of data sets derived from the GOES-8 5-channel Imager and the DMSP passive microwave instrument suite (SSM/I, SSM/T, SSM/T2) have been acquired for the Gulf of Mexico-Caribbean Sea basin. Whereas the methodology is being tested over this basin, the algorithm is being designed for portability to any open-ocean region.

The algorithm design takes advantage of the high temporal resolution of the GOES-8 measurements as well as the physical relationship between the SSM measurements and water vapor, cloud liquid water, and rainfall. Used together, these measurements can be used to retrieve the geophysical parameters in the water budget equation.

Figure 1. Region of study and station upper air network.

The region of study (Fig. 1) consists of the Gulf of Mexico and Caribbean Sea Basin. This region is surrounded by a network of upper air stations (shown in Fig. 1) providing an independent estimate of the water vapor transport into the region. The study extends beyond an annual cycle, consisting of six months: October-97, January-98, April-98, July-98, October-98, and January-99. A main goal of the study is to

document the daily and seasonal properties of all theterms in the atmospheric water budget equation over a complete annual cycle.

2. SCIENTIFIC OBJECTIVES

Past research on the atmospheric water budget has concentrated its efforts in studying the seasonal and annual properties of the water budget on both regional as well as global scales. In doing so, a simplified form of the water budget equation that consists of a balance between evaporation minus precipitation (E - P) and the divergence of water vapor transport () has been used (Benton and Estoque, 1954; Starr et al., 1958; Hastenrath, 1966; Rasmusson, 1967, 1971; Starr and Peixoto 1971; Peixoto and Oort, 1983; Chen and Pfaendtner, 1993). In this study, such a balance will not be assumed. Instead, the scientific objectives of this research are focused on testing the following hypothesis: local changes of storage of precipitable water and condensates within convectively active regions are significant and should be considered in space-time restricted water budget calculations. Thus, the conventional time-averaged form of the water budget equation used in previous studies, consisting of the balance mentioned above, does not generally retain its validity when the budget calculations are made seasonally or regionally, particularly over convectively active regimes.

With this in mind, the scientific objectives of this research are twofold: (1) to develop a purely satellite-based retrieval methodology, using multi-spectral measurements from GOES-8, SSM/I, SSM/T, and SSM/T2, to calculate the atmospheric water budget over the Gulf of Mexico-Caribbean Sea Basin. This includes the retrieval of water vapor content (W), cloud liquid water content (Wc), divergence of water vapor transport , precipitation (P) and surface evaporation (E); (2) to quantify the uncertainty in a convectively active tropical-subtropical region stemming from the assumption that the local rate of change of precipitable water and cloud liquid water are negligible in the context of the regional-seasonal atmospheric water balance.

3. METHODOLOGY

The regionally averaged atmospheric water balance equation is given by:

where the brackets represent a regional average, W and Wc are the vertically integrated water vapor and cloud liquid water, Q is the vertically integrated water vapor transport , E is the surface evaporation, and P is precipitation. The first two terms on the LHS of this equation represent the time rate of change of the vertically integrated water vapor (W) and cloud liquid water (Wc), while the third term represents the divergence of the water vapor transport. In deriving this equation it has been assumed that the horizontal transport of condensates is small compared to that of water vapor because circulations within convectively active systems mainly take place in the vertical.

The methodology consists of retrieving the precipitation, surface evaporation, and vapor-cloud water storage terms from a combination of satellite techniques using high frequency GOES-8 measurements and SSM/I, SSM/T, and SSM/T2 microwave measurements. The water vapor advection term is then obtained as a residual from the balance equation. This leads to a purely satellite-based method for deriving the full set of terms required for the atmospheric water budget equation without requiring information on the wind velocity profile.

The algorithm will be validated by comparing the divergence of the water vapor transport term diagnosed from the satellite algorithm, with that obtained from a network of land-based upper air stations that uniformly surround the basin (Fig. 1). To calculate the divergence term from the ring of upper air stations, we can apply Green's theorem to obtain the moisture flux divergence from the line integral along the boundaries of the basin, that is:

This type of line integral technique has been successfully applied to calculate vertically integrated moisture budgets (e.g., over the Gulf of Carpentaria in Australia by McBride et al., 1989). We also intend to validate the rainfall estimates using the combined radar-radiometer facility algorithm from the Tropical Rainfall Measuring Mission (TRMM).

Eventually, we intend to analyze the time series of regional averages and time averaged regional maps for each of the terms in the balance equation on daily and monthly time scales. Regional averages will be obtained for the entire Gulf-Caribbean basin and for the Gulf and Caribbean, separately. This provides a means for identifying differences between the Gulf of Mexico and Caribbean Sea regional water budgets. The calculations will be done for total columar amounts and for a three layer profile stratified into the 1000-700 mb, 700-400 mb, and 400 mb-tropopause layers.

4. PRELIMINARY RESULTS

Preliminary budget results will be presented, obtained from combining an integrated set of algorithms for the retrieval of the different terms in the water balance equation. Future research will include optimizing the selection of algorithm modules needed to obtain the time-evolution of an open-ocean atmospheric water budget and conducting a detailed analysis of an annual cycle over the Gulf of Mexico-Caribbean Sea region.

5. SUMMARY

This brief paper is about a work in progress. The development of a portable, purely satellite based algorithm for the retrieval of the atmospheric water budget over the Gulf of Mexico/Caribbean Sea basin is presented. The water budget calculations are made using a combination of multispectral measurements derived from the GOES-8, SSM/I, SSM/T, and SSM/T2 instruments. The results of the research will test whether ignoring the water vapor and cloud liquid water terms in the water balance equation is a valid assumption in estimating water budget processes at regional-seasonal space-time scales, particularly for convectively active environments. It will also test if retrieving the water budget purely from satellite observations, without requiring a wind velocity profile, is a suitable approach by comparing the residue transport term obtained from the retrieval algorithm to conventional calculations obtained from upper air soundings surrounding the Gulf-Caribbean basin.

6. REFERENCES