While quite a few of the NPS visitation numbers are obviously "soft", some seem considerably inflated.  For example, I am quite familar with both parks and am unable to believe that Olympic's annual visitation has actually been almost triple that of Mount Rainier in recent years.  The parking lots at Rainier are far larger and I have never waited in a mile-long line to enter Olympic.  Most Olympic attractions are also more than twice the driving time from local population centers in 'Pugetopolis' as Rainier. 

Both park's websites used to have links explaining their counting methodology, but I can no longer find them.  Both used traffic counter figures multiplied by a persons/vehicle factor, which was then decreased by a percentage of 'Non-Recreational' visits.  I think OLYM considerably undercounts non-recreational traffic where the US 101 loop passes through the park at Lake Crescent and the Kalaloch strip.  They probably also count the same visitor vehicles multiple times as they turn up each spur road like Hurricane, Soleduc, and the Hoh.  This may also occur at Rainier when a vehicle visits Paradise and Sunrise on the same day.

How do they measure "visitation" on the GW Parkway?  This is a major commuter thoroughfare with the vast majority of traffic visiting nothing. Are those commuters being counted as park visits?

Wasn't there something written about this a while back?  One of the issues is the difficulty of keeping track, and then there's the issue of what constitutes a "visit" (does driving through count as a visit, or do you have to do more?).  Because the parks are so unique, I think the parks have leeway in determining what constitutes a visit, but one question might be if someone leaves and comes back, how many visits is that?  For keeping track, I think in some cases they have to estimate.  But there were also issues with things where they did have a way of keeping track.  For example, I remember something where a counter of some kind had its battery run out, and so they had no way of knowing and therefore had to estimate or extrapolate somehow.  I think there's a place where each park says how it estimates visitation, and that can provide more information or insight about where the stats come from.

In the Smokies, Newfound Gap road is the most direct thoroughfare to NC so if someone drives from state to state, they are considered a National Park visitor. Therefore, if casino trafiic in Cherokee is up, then the Smokies gets to count it and subsequently ask the public for more money to subsidize the casino traffic or Dollywood splash country bonanza coming the other way.  The majority of Smokies "visitors" never leave their vehicles and hardly even are aware they are in a National Park.  Yet they claim 9 million.   Just a numbers game, the proof lies in the fact that of that 9 million a paltry 69000 ever visit the backcountry in a given year.  Yet the bulk of monies to the Smokies goes to roads and visitor centers along NF gap road etc. 

Visitation figures for many areas are a very complicated question, especially for areas that have many possible entry points. Short of having someone physically count each person/vehicle entering the park (which is rarely feasible 24/7, even for areas that have an entrance station), at best all that can be hoped for are reasonable estimates.

I'm guessing there's still a small staff in an NPS "statistical" office that oversees this effort. Twice during my career, I was working at a park when one or two of them spent several days in the area, doing spot checks and surveys to try to update formulas specific for that park, such as average # of people per vehicle, how many were locals just "passing through" vs. "recreational visitors," etc. 

To a large degee, perhaps these numbers shouldn't matter.  Sure, a Yosemite or Grand Canyon clearly needs more funding than a small historical site, but these numbers (and changes up or down from year to year for a park) probably receive more attention than they deserve by those controlling the purse strings to decide whether a park (or the agency as a whole) deserves more or less funding. 

In my humble opinion, such variations + or (-) in visitation are only important if they indicage a major change in use of an area due to some significant factor, either internal or external to the park. 

John, once again, you spew facts like you pretend to have these absolute answers, when they are quite off.  Over a million go up to Clingmans Dome every year, About a 1/3rd of a million go further than a 1/2 mile in the park, and around 2 million + go into Cades Cove every year.  In fact Cades Cove would be in the top 10 of most visited spots in the National Park system.  By the way, they do keep traffic counters and those stats are online too, and each time a car passes by those counters, it's clicked, and those hard numbers are out there.  But, hey... do I need to show you where those are to? Since you seem to know everything there is to know about what's happening in the Smokies, i'm sure you already know these numbers, right?

ec,

GW Memorial Parkway is the admistrator for several small park units like Great Falls Park, Ft Hunt Park, Gravelly Point, Fort Marcy, Marine Corp Memorial etc.  Looks like the traffic counters for recreation visits are on entrance/access roads from the GW Pkwy to those areas.

Jim Burnett does a good job of summarizing the difficulties of estimating NPS visitation.  I agree that short-term flucuations in these numbers should not be used to justify budget reductions (or increases, for that matter) for specific units.  However, I believe those sorts of logistical difficulties and fiscal pressures he describes introduce an overall bias towards overcounting.  IMO, these "soft" and inflated visitation numbers are also a major factor in overestimating the economic impacts of NPS units.