Quantifying the utility of enhanced meteorological observations in the New York City metropolitan area with an Observing System Simulation Experiment

As part of the Department of Homeland Security's Urban Dispersion Program, a permanent meteorological network consisting of six sodars will be deployed in the New York City metropolitan area. In the event of the atmospheric release of hazardous material, data from this network can be used with dispersion models to advise emergency responders on the transport and diffusion of such hazardous material. Because of the variety of weather phenomena endemic to the NYC area, such a network must be able to resolve wind fields in circumstances ranging from sea breezes from the south-east, frontal passages from the north, intense convection, etc.

To justify the investment in such a network, an observing-system simulation experiment (OSSE) has been carried out to illustrate the ability of the network to provide adequate data on wind speed and wind direction in 36 randomly-chosen synoptic situations over the course of one year (2003). This OSSE compares a “truth” wind field to wind fields generated by assimilating data from several locations, including those comprising the observing network locations. The fidelity of the assimilated wind fields to the “truth” field provides a measure of the utility of the network.

For this OSSE, the “truth” field for each of the 36 days is represented by a 24-hr forecast using the Navy's COAMPS mesoscale model. Only data from the innermost of 5 nests, centered on Manhattan and run at 400m horizontal resolution is utilized. The assimilated fields are generated by extracting pseudo-sodar observations (at the chosen sodar siting locations as well as several other candidate locations) from that innermost COAMPS nest, adding in random error, and then assimilating those observations in a diagnostic wind-field model, ADAPT, which was developed for use at the National Atmospheric Release Advisory Center located at Lawrence Livermore National Laboratory, Livermore, California. The RMS error between the assimilated fields and the “truth” wind field is quantified, allowing a ranking of the sensor locations. We will present the rankings of the sensor locations, including recommendations for locations of subsequent sensor deployments. UCRL-ABS-214020