Author(s)

J.C. Ruiz-Suarez, O.A. Mayora, R. Smith-Perez & L.G. Ruiz-Suarez

Abstract

A neural network-based prediction model of ozone for Mexico City J.C. Ruiz-Suarez", O.A. Mayora", R. Smith-Perez", L.G. Ruiz-Suarez^ "Institute* Tecnologico y de Estudios Superiores de Monterrey, Campus-Morelos, Apdo. Postal 99- C, Cuernavaca, Morelos 62050, Mexico ^Centro de Ciencias de la Atmosfera, Universidad Nacional A. de Mexico, Circuito Exterior, Cd. Univer sit aria, Mexico D.F. 04510, Mexico ABSTRACT This paper describes current work for developing a short-term forecasting mod- el for ozone in Mexico City. The structure of the model is based on a recently created paradigm, the Holographic Associative Memory (HAM). The HAM is able to store a large amount of daily patterns and address any of them in a effi- cient way. Preliminary results of ozone forecasting are reported and some important conclusions are drawn. INTRODUCTION Air quality in Mexico City is an important public concern. Average daily emis- sions of primary pollutants, such as hydrocarbons, nitrogen oxid

Keywords