The modular ocean data assimilation system (MODAS) was designed to meet the U.S. Navy’s need to produce rapid estimates of present and near-term ocean conditions, often in situations where little or no in situ data are available. One of the problems facing oceanographers is the scarcity of in situ measurements to use in estimating the temperature and salinity structure of the ocean. Traditionally, we have resorted to climatologies derived from all the available historical data (Teague et al., 1990), or performed our own measurements using instruments such as expendable bathythermographs, which can be quite expensive and time-consuming. Such measurements may also be impractical in the case of mission planning or emergencies. In Navy mission execution, operators typically have access to in situ point measurements but have not had a means to fuse these data into a single picture of the ocean environment, or a means to blend these data with climatologies to fill in the data voids. Fortunately, there are several satellites that make routine measurements of sea surface temperature and height anomalies from space. These measurements can reveal information about the locations of surface features such as fronts and eddies, as well as subsurface features that can significantly affect acoustic propagation (Carnes and Mitchell, 1990; Carnes et al., 1994). The combination of in situ measurements and remotely sensed temperatures and heights forms a complex sampling pattern. The goal of MODAS is to combine these disparate data types and irregular sampling patterns to form a single integrated analysis of temperature and salinity on a regular grid.