Although a wide variety of transportation data sets involve discrete values scattered across space and time, there are currently few techniques for properly analyzing such data. This paper describes a new dynamic spatial ordered probit (DSOP) model and demonstrates the model's use in a case of ozone concentration categories. With outputs of photochemical models for the Austin, Texas, region from a 24-h period, model parameters are estimated with Bayesian techniques. The results illuminate key relationships, many intuitive but generally obscured by complex upstream model systems. Through the use of 132 4- x 4-km surface grid cells as observational units, values that exhibit strong patterns of temporal autocorrelation but appear strikingly random in a spatial context (after controlling for local land cover, transportation, and temperature conditions) are found. Although transportation and land cover conditions appear to influence ozone levels, their effects are neither as instantaneous nor as practically significant as the influence of temperature. The DSOP model proposed in this paper is able to accommodate the unusual dynamics and spatial evolution of the ordered response categories that are inherent in ozone concentration data.