Improving the resolution of gridded-hourly mobile emissions: incorporatingspatial variability and handling missing data

To more accurately predict hourly running stabilized link volumes for emiss ions modeling, a new method was recently developed that disaggregates the p eriod-based model link volumes into hourly volumes using observed traffic c ount data and multivariate multiple regression (MMR). This paper extends th e MMR methodology with clustering and classification analyses to account fo r spatial variability and to accommodate model links that do not have match ing observed traffic count data. The methodology was applied to data collec ted in the South Air Basin. The spatial analysis resulted in identifying fi ve clusters (or 24-h profiles) for San Diego and two clusters for Los Angel es. The MMR models were then estimated with and without clustering. For San Diego, the disaggregated model volumes with clustering were much closer to the observed volumes than those without clustering, with the exception of the a.m. period. For most hours in Los Angeles, the predicted volumes with clustering were only slightly closer to the observed volumes than those pre dicted without clustering, suggesting that spatial effects are minimal in L os Angeles (i.e., that 24-h volume profiles are fairly similar throughout t he region) and clustering is not necessary. Finally, two classification mod els, one for San Diego and one for Los Angeles were developed and tested fo r network link data that does not have matching observed count data. The re sults indicate the procedure is relatively good at predicting a cluster ass ignment for the unmatched location for Los Angeles but less accurate for Sa n Diego, (C) 2001 Published by Elsevier Science Ltd.