Updated: Jun 24, 2020
The TSA charts are update here.
NOTE: Forecasts are hypothetical and speculative.
Forecasts with outlook beyond 30 days are extremely speculative.
Peruse cautiously, and use responsibly.
Disclaimer: "I'm with the tour group."
As you can see in our "since bottom" chart, we usually add our best guess for the doubling of traffic from the previous doubling. For the last few weeks, we were expecting the doubling to occur on or about July 7th, 2020.
This week's forecast shows that the acceleration we saw last week slowed, but the doubling of traffic from 350,000 to 700,000 should still occur several days sooner than predicted by the early June forecasts - sometime between June 25, 2020 and July 7, 2020.
This week's data also suggests that we should cross the important milestone of 1,000,000 sometime at the end of July or the first half of August, 2020.
The forecast for July looks rather linear (suspect), so we're still rooting for 1,000,000 daily passengers sometime in July, 2020. The polynomial for the traffic data still shows the 1,000,000 mark around the last week of July, 2020.
The poly charts are explained in detail in this blog post.
The poly suggests that new traffic data arriving matches the previous patterns and matches the last few weeks' "forecast", or poly trend-line, for the rest of the year. It does so in such a way that the poly curve fit continues to increase in fitment over previous weeks - from 88.8% to 93.24% this week.
The percentage line (red) seems to mirror the actual traffic data, however, this is only an approximation since each percentage normalizes to the traffic number for 2019. On a short term poly (30-60 days), the approximation is close enough to be somewhat useful - the percentage scale on the right roughly correlates to the traffic scale on the left for the latest data point (June 21, 2020). The 30 and 60 day extrapolations are an even rougher estimate of that correlation, but they provide a general trend.
On a 120 poly, one should really separate the polys for each dataset (one for traffic, and one for the percentages). Microsoft won't let me do that easily. I plan to find the time to add it shortly, in the same graph or separately.
Basically, there are two charts combined into one: the traffic chart with the polynomial curve (poly) and the percentage line, which normalizes to 2019 traffic. Up to the chart date, June 21, 2020, the two charts seemingly coincide - the traffic scale and the percentage scale roughly match the current values. For future dates, they diverge.
Therefore, you have to consider and keep in mind the meaning of the "100%" for the percentage value axis, or scale:
100% would be referencing the average traffic number for 2019, for the period covering the available traffic data since traffic bottom in 2020, one year ago.
In other words, the period from 4/14/2019 - 6/22/2019:
Traffic bottom: 4/14/2020.
Chart Date: 6/21/2020.
Days in 2020 since traffic bottom: 68.
Average total traveler throughput, one year ago, same period of data availability in 2020 (4/14/2019 - 6/22/2019): 2,433,228.
So that 100% on the far right should reference an average traffic throughput of 2,433,228 daily passengers for the period of 4/14/2019 - 6/22/2019. As you can see, in the graph below, it is "across" the 2,750,000 daily passenger number - but it does not relate to that number.
The 2,433,228 daily passenger number is roughly "across" the 90% scale - but they are not correlated either. Compare the polys for the traffic data and the percentage data.
This shows the divergence mentioned earlier, which looks to be about 10% in October.
This average number of 2,433,228 representing the 100% in the percentage scale may not be representative of the usual October numbers for last year or the last few years. It's a very, very rough approximation. We would need TSA October 2019 data, and I've not seen it anywhere. I did find "TSA Year In Review: 2019," which highlights July 4th and December 1, 2019 traffic numbers.
You cannot read too much into all of this, as a 120 day poly is outrageously too far ahead of the data - but perhaps we can deduce this:
This approximation, plus the poly for the available traffic data and the poly for the percentage data, indicate the possibility of about 90% traffic vs 2019, in the last quarter of 2020 - only if future data continues to match the same pattern as the current data pattern since traffic bottom on 4/14/2020.
As someone cleverly pointed out ...