### Table I. IAH Passenger Departure and Arrival Statistics

### Table 2: Daily arrival and departure rates.

"... In PAGE 4: ...2 Daily average Given that the arrival and departure rates that we com- puted in the previous section are normalized, we need the actual number of arrivals and departures to compute the hourly value. Table2... In PAGE 5: ... This implies that the number of visits at APs did not increase much during the two months of the traces. To model the transitions between APs, the average over APs presented in Table2 may not be enough; al- though APs within a cluster follow similar hourly vari- ations, they are unlikely to have similar numbers of ar- rivals. Thus, we need to consider what kind of distri- bution the daily number of arrivals follows within each cluster.... ..."

### Table 13: Comparison of annual departures for GAES-Move and STATFOR data

2004

"... In PAGE 40: ...3.2 Comparison with STATFOR Statistics Table13 hereafter shows the comparison of the GAES-Move European data with the domestic and international traffic statistics available in the EUROCONTROL STATFOR database. Note that Belgium and Luxembourg data are combined in STATFOR, therefore the corresponding GAES-Move statistics were recalculated.... ..."

### Table 3-l. Single Vehicle Roadway Departure Causal Factor Analysis

in Examination

"... In PAGE 19: ...The statistics shown in Table3 -l represent the causal factor percentages of the total SVRD case sample. No single causal factor predominates.... ..."

### Table 3: Departure times deviations and comparison of means test

in FOR THE

2006

"... In PAGE 12: ...able 2: Time of peak for different regions..................................29 Table3 : Departure times deviations and comparison of means test.... In PAGE 35: ... The last column gives the number of travelers during the peak-period and the column to its left gives the percentage of these travelers making their trips in the observed peak 5-minute window. Table3 summarizes the results for trips to the left and to the right of the observed peak window. The table gives the mean of deviations of trip departure times from the observed time of peak.... In PAGE 40: ...rom the dataset. Of the total 5,572 work and work-related trips during the day 6.7 percent occurred during the 5-minute interval between 7:25 AM and 7:30 AM. Discussion Using difference of means test ( Table3 ) we find that the z-statistic is greater than (+1.96) or less than (-1.... In PAGE 41: ... However this does not give us conclusive evidence about peak-spreading, as in 2 of 5 regions, the left and right deviations are moving in the same direction, and in only 1 significant case is the right deviation moving away from the peak (and opposite in sign to left deviation). From Table3 , we can also see that the difference in the mean deviations carries a negative sign in 9 out of the 12 cases, except for the right side trips in Atlanta, St. Louis and Twin Cities.... In PAGE 41: ... Louis and Twin Cities. Table3 shows that the mean left side deviations for the recent years are consistently greater in magnitude than the previous year. This implies that the trip departure times for the trips to the left are shifting further away from the peak to the left.... ..."

### Table 5: Arrival and departure rates as a function of time of day

"... In PAGE 4: ... Khoury, Kamat and Ioannou The aircraft mix for this simulation model is 33% heavy, 33% medium size and 33% light, but can be easily modified in the model to reflect any other aircraft mix. The DTW arrival and departure schedule shown in Table5 was obtained from the BTS database by looking up the DETAILED STATISTICS for arrivals and departures according to each airline and then computing the daily averages for the busy month of December 2004. The purpose of this simulation is to determine the expected average and maximum daily waiting times for arriving and departing airplanes for any runway configuration.... In PAGE 4: ... The purpose of this simulation is to determine the expected average and maximum daily waiting times for arriving and departing airplanes for any runway configuration. This information is used to determine if the airport is capable of supporting the demands shown in Table5 with acceptable delays. Other assumptions were taken into consideration.... In PAGE 4: ... The first portion of the network comprised of the PArrival Combi activity; the ArvlSchdlrs Queue; and the AS1, AS2, and AP1 links models the airplanes arrival to the system. Each of the twenty four resources of type ArvSchd, which are initially in ArvlSchdlrs queue, represents the information in the first two columns of the corresponding row in Table5 defined earlier, and enables the creation of a 2032 Table 5: Arrival and departure rates as a function of time of day ... In PAGE 5: ...ArvSchdlr.NextArv apos;; SampledIAT is the interarrival time between airplanes and is based on the rates from Table5 . StartHr and EndHr define one hour range timing for each arrival scheduler resource and they are giving in Table 5.... In PAGE 5: ...edIAT: PArrival.ArvSchdlr.NextArv apos;; SampledIAT is the interarrival time between airplanes and is based on the rates from Table 5. StartHr and EndHr define one hour range timing for each arrival scheduler resource and they are giving in Table5 . The variable CurHour and the property NextArv, which is referenced in the above expression, are calculated as follows: VARIABLE CurHour Mod[SimTime/3600+SimBeginHour,24]; VARPROP ArvSchdlr NextArv apos;SampledIAT+CurHour*3600 lt;EndHr*3600 ? SampledIAT: SampledIAT+(24-(EndHr- StartHr))*3600 apos;; When PArrival terminates, the arrival scheduler is released to ArvlSchdlrs, creating by this another instance of PArrival.... In PAGE 6: ... This network illustrates two major portions: the first one introduces departing airplanes to the system through arrival scheduler resource and the second models departures of airplanes through resource Plane. Airplanes departing are generated independently of arrivals according to the rates specified in Table5 and are placed in the PlanesWaitD queue. Departures are modeled by the portion of the network comprised of the PDeparture Combi activity, the DptSchdlrs Queue, and the DS1, DS2, and DP1 links in much the same way as the arrivals to the system are modeled.... ..."

### Table B-7. Binary Logit Model Estimation Results of Departure Time Change Because of the Weekend Closure of I-405

1998

### Table 6 CEO Departure Rates by Levels of CEO Ownership Ownership levels are based on classifications used in Dahya et al (1998) Sample is 2110 All Share firms

"... In PAGE 20: ...5% post-Cadbury. Table 5 about here Table6 partitions the sample according to the CEO ownership classifications used by Dahya et al (1998). For CEO ownership levels of less than 1% our forced departure rates are similar, at over 4%.... In PAGE 21: ... Table6 about here 5.5 Summary The univariate tests in this section provide useful indicators of associations between firm performance, governance variables, regulatory regime and non-routine CEO departures in UK All Share firms in the first half of the 1990s.... ..."

### Table 3. Projected 2005 Delay Statistics

"... In PAGE 34: ...Because the CBM requires input in the form of changes in block time, the next step was to convert the figures from Table3 to percent changes in block time. The conversion requires an assumption about the average block time for departing and arriving flights at each airport.... ..."