No abstract

Controlling fugitive emissions from leaks in petrochemical industry process equipment now requires periodic monitoring of valves, flanges, pumps etc., typically on a quarterly basis. Previous studies have shown that over

Multiple regression techniques were used to predict surface shelter temperatures based on the time period 1986–89 using upper-air data from the European Centre for Medium-Range Weather Forecasts to represent the background climate and site-specific data to represent the local landscape. Global monthly mean temperature models were developed using data from over 5000 stations available in the Global Historical Climate Network (GHCN). Monthly maximum, mean, and minimum temperature models for the United States were also developed using data from over 1000 stations available in the U.S. Cooperative (COOP) Network and comparative monthly mean temperature models were developed using over 1150 U.S. stations in the GHCN. Initial correlation analyses revealed that data from 700 mb were sufficient to represent the upper-air or background climate. Three-, six-, and full-variable models were developed for comparative purposes. Inferences about the variables selected for the various models were easier for the GHCN models, which displayed month-to-month consistency in which variables were selected, than for the COOP models, which were assigned a different list of variables for nearly every month. These and other results suggest that global calibration is preferred because data from the global spectrum of physical processes that control surface temperatures are incorporated in a global model. All of the models that were developed in this study validated relatively well, especially the global models. Recalibration of the models with validation data resulted in only slightly poorer regression statistics, indicating that the calibration list of variables was valid. Predictions using data from the validation dataset in the calibrated equation were better for the GHCN models, and the globally calibrated GHCN models generally provided better U.S. predictions than the U.S.-calibrated COOP models. Overall, the GHCN and COOP models explained approximately 64%–95% of the total variance of surface shelter temperatures, depending on the month and the number of model variables. The R2's for the GHCN models ranged between 0,86 and 0.95, whereas the R2's for the COOP models ranged between 0.64 and 0.92. In addition, root-mean-square errors (rmse's) were over 3°C for GHCN models and over 2°C for COOP models for winter months, and near 2°C for GHCN models and near 1.5°C for COOP models for summer months. The results of this study—a large amount of explained variance and a relatively small rmse—indicate the usefulness of these models for predicting surface temperatures. Urban landscape data are incorporated into these models in Part II of this study to estimate the urban bias of surface ternperatures.

A methodology is presented for estimating the urban bias of surface shelter temperatures due to the effect of the urban heat island. Multiple regression techniques were used to predict surface shelter temperatures based on the time period 1986–89 using upper-air data from the European Centre for Medium-Range Weather Forecasts to represent the background climate, site-specific data to represent the local landscape, and satellite-derived data—the normalized difference vegetation index (NDVI) and the Defense Meteorological Satellite Program (DMSP) nighttime brightness data—to represent the urban and rural landscape. Local NDVI and DMSP values were calculated for each station using the mean NDVI and DMSP values from a 3 km × 3 km area centered over the given station. Regional NDVI and DMSP values were calculated to represent a typical rural value for each station using the mean NDVI and DMSP values from a 1° × 1° latitude–longitude area in which the given station was located. Models for the United States were then developed for monthly maximum, mean, and minimum temperatures using data from over 1000 stations in the U.S. Cooperative Network and for monthly mean temperatures with data from over 1150 stations in the Global Historical Climate Network. Local biases, or the differences between the model predictions using the observed NDVI and DMSP values, and the predictions using the background regional values were calculated and compared with the results of other research. The local or urban bias of U.S. temperatures, as derived from all U.S. stations (urban and rural) used in the models, averaged near 0.40°C for monthly minimum temperatures, near 0.25°C for monthly mean temperatures, and near 0.10°C for monthly maximum temperatures. The biases of monthly minimum temperatures for individual stations ranged from near −1.1°C for rural stations to 2.4°C for stations from the largest urban areas. There are some regions of the United States where a regional NDVI value based on a 1° × 1° latitude–longitude area will not represent a typical “rural” NDVI value for the given region, Thus, for some regions of the United States, the urban bias of this study may underestimate the actual current urban bias. The results of this study indicate minimal problems for global application once global NDVI and DMSP data become available. It is anticipated that results from global application will provide insights into the urban bias of the global temperature record.

A cost-effective fugitive emission reduction program should focus on locating and repairing the very high leakers. Although these components (such as valves, pumps, compressors, flanges, etc.)