If the images from the Tahoe Donner firewatch camera are reliable, we can virtually prove that Kiely Rodni’s CR-V did not enter Prosser Reservoir on the night of the Prosser party. This is assuming that Kiely was there to begin with – but that is another matter altogether. So let’s go through the steps I took to get to this conclusion.
Eyes in the rock that spy Eyes in the pine-tree dark
Euripides, The Bacchae
Not long after Kiely Rodni’s CR-V was discovered in Prosser Reservoir by Doug Bishop, Nick Rinn, and the rest of the AWP team, many people began to wonder if the car was placed there after the official searches had concluded. There were at least three agencies, including CHP and the FBI, who searched the reservoir up and down, multiple times. They used all the right tools, and Kiely’s car was discovered in an area which wasn’t that deep. If the car was there, it would have been found. The only sensible conclusion we can make is that Kiely’s car was not in Prosser Reservoir until after August the 17th.
Steve Fischer, a P.I., shot aerial drone footage on Saturday, August 20th. The car may be visible in this footage. If so, this means that the car was probably put into Prosser Reservoir between the 17th and the 20th:
I looked through footage I shot Saturday afternoon at Prosser. Footage I had not uploaded yet. To think i was just feet from her. Thankfully AWP would find her 18 hours later #kielyrodnipic.twitter.com/aC8zNfQYpA
— SF INVESTIGATES • STEVE FISCHER (@SF_investigates) August 25, 2022
Last year, I had an idea about the images from the Alder Hill firewatch camera. I suspected that the nighttime frames did not correctly align with the daytime frames. This has serious implications, because so many people dropped this case due to the Alder Hill images. The conclusion among many commentators was that Alder Hill proved that Kiely’s car entered the reservoir at 00:33 on August 6th. And they put the story to bed.
If this imagery was improperly interpreted, this should be rectified as soon as possible. I did notice a small alignment issue between day and night images, and I did some preliminary work on trying to align them. What I was seeing showed that the object that seemed like Kiely’s car was someone else’s car. I did show this to a few people privately, and they all agreed. However, I noticed a mistake, and I decided to publish it later, after I had fixed the issues. But, I don’t think it’s relevant anymore, because the Tahoe Donner images seem to be much more reliable, as I will show.
Some quick background: some months ago, I had downloaded a bunch of frames from the Tahoe Donner firewatch camera. I did not get the memo about this camera before then. I then proceeded to completely forget about them until last week.
In any case, I decided to check these images, and I overlaid a nighttime frame over a daytime frame. There is a little bit of frame float with these Axis cameras, probably due to the camera mounting mechanism. But the alignment seems pretty good otherwise.
There is one problem: these Axis cameras are not properly set up. When the light levels get too low, the camera has to switch from colour to b&w. This is by design, and if you look through the frames, which are separated by two seconds each, you can pinpoint the exact moment when this switch occurs. The problem is that when light levels get too low, there is not enough contrast in the scene, and the camera doesn’t know what to focus on. So it defaults to closest focus instead of infinity.
This AF problem is very annoying because it makes it difficult to precisely align any two frames. Not to mention the fact that light sources become blobs instead of points. That matters, because if a light source is too faint to form a blob, it won’t be picked up by the camera’s sensor. Despite all this, I have managed to find something meaningful.
I won’t focus on the Alder Hill camera too much. I wrote a blog post about that already. But just for context, here is, apparently, the last few seconds of light from Kiely’s CR-V in Prosser Reservoir, according to Alder Hill:
The last moments of Kiely’s car (supposedly), overlaid with an image from earlier that day
(Although not directly relevant to this post, note that a car’s electrics can last for much longer than a couple of minutes submerged).
It was this series of images which led to so many people declaring Kiely’s case as ‘closed’. Again: in my opinion, and the opinion of many others, Kiely’s car was not in the reservoir until after August 17th, when the searches stopped. This is very much a leap of faith, because I personally do not believe that the Alder Hill frames are manipulated. I just think that there is an alignment issue. Perhaps it’s partly due to the wind; or perhaps that unit has a lens which changes perspective between near focus and infinity focus.
Let’s now look at a daytime frame from the Tahoe Donner camera. Like many cameras in this network, it is an Axis model Q6075-E. Here is the view of Prosser Reservoir at 17:00:
Here is the same view, but at night, supposedly just as Kiely’s car is sinking to the bottom of the reservoir:
How good is the alignment between these two frames, over seven hours apart? Let’s overlay them and see:
These would be much easier to align if the camera had retained infinity focus. Anyway, as far as I can tell, the alignment is very close. Good enough to try and draw a conclusion? That is up to you to decide. So now we will put a crosshair over the defocused light in the night frame, which is supposedly Kiely’s car. We will be as precise as we can, given the vague nature of defocused highlights:
Now, we combine the crosshair frame and the daytime frame:
Now we remove the nighttime frame, and put a little dot in the centre of the crosshair, and overlay that onto the daytime frame:
Now, a crop from the region we might call the ‘little peninsula’, the last point that Kiely’s car supposedly touched land:
If we sharpen and saturate the image, as best we can, it’s evident that the green dot sits on land, not water:
To the best of my knowledge, and to the best of my visual perception, that green dot, which is the centre of the light blob, which was supposedly the point at which Kiely’s car sank, appears right at the tip of the little peninsula. Perhaps due to alignment issues, it’s a little bit further back from the water. But it appears to be on land, not on water.
If all of this is accurate, and correctly done, then we can conclude that we have virtual proof that Kiely’s car did not enter Prosser Reservoir on the night of the Prosser party. And it follows that Kiely’s car must have been dumped into the reservoir after police searches ended, but before AWP arrived. This evidence matches very well with the fact that the CHP, FBI and local police did not find any car in Prosser. If it was there, they would have found it. The only question is why the Alder Hill images do not tell quite the same story.
“I’ve got the front desk now. He was never here.”
The questions that remain: how did Kiely die, when did she die, where did she die, and where was her body kept? We have no evidence that she was at the Prosser party. It is further concerning that we have no evidence that Kiely was alive before the morning of Friday the 5th.
Bayes’s Theorem (BT) is a statistical equation that helps determine the probability of a hypothesis or explanation. You can apply BT to almost anything: history, science, medicine, law, criminology, cryptology, etc.
If you saw a UFO, what are the chances that it was an alien spacecraft? What are the chances that you actually have breast cancer, based on the data from a mammogram (albeit a now obsolete test)? What are the chances that a criminal suspect committed the crime? What are the chances that a police detective will order an espresso at the local café?
Like anything in statistics, the output can only be as good as the input. Here, I want to do a quick and simple Bayesian analysis to determine whether or not Kiely Rodni died from the result of foul play.
Bayes’s Theorem (BT) includes two main variables, and two supplementary variables, all between 0.00 and 1.00. 0.00 is 0% and 1.00 is 100%. The easiest variable to understand is prior probability. That is, in general, how likely is is that your hypothesis is true, based on previous experience? In the case of females found dead in submerged vehicles, we do have fairly good prior probability: 66%, properly written as 0.66. SAR diver, Nick Rinn, explains it:
Nick Rinn: “about 2/3 of those (women found dead in submerged cars) end up being homicide or foul play.”
Before I go any further, here is a basic illustration of Bayes’s Theorem – it’s actually not difficult to understand:
Adapted from Proving History: Bayes’s Theorem and the Quest for the Historical Jesus by Dr. Richard Carrier
And here is the formal way to write it:
And here is the basic form, just in case you don’t notice it in the above versions:
It’s basically asking this question: out of all possible outcomes, how true can your hypothesis be, given the evidence you have for it?
Just as a very simplified example, let’s say that there is a room filled with 100 computers. I own five of these computers, and therefore, other people own 95 of those computers. So, if you randomly selected one computer, what are the chances that it’s mine? You already know the answer, which is 5%: 5/(5+95) = 5/100 = 0.05, or 5%.
So, you have a 5% chance of randomly choosing a computer in that room that belongs to me. To be honest, that was an overly simplified application of BT. But it showed, hopefully clearly, the function of the equation, and why A appears both above and below the line.
Bayes’s Theorem works with two main variables: the chances of your hypothesis (or explanation) being true, and the chances of it not being true, both based on background information. A shorter way to put it: your explanation versus the total of all other explanations. Logically, they both must add up to 100%, or 1.00.
Each main variable has a supplementary variable. The function of the supplementary variable is to modify to main one. Each main variable has a maximum value, and this can be reduced depending on the strength of the evidence for it.
In our case, based purely on background information, there’s a 66% chance that Kiely died by means of foul play. That’s prior probability. But wait a minute! We have evidence specific to Kiely’s case. That specific evidence cannot increase the prior probability, but it can reduce it. This specific evidence can be given a value between 0.00 and 1.00. So, the figure of 66%, or 0.66, can be as low as zero if none of the evidence supports it (0.66 x 0.00), but no higher than itself if all the evidence supports it (0.66 x 1.00).
In the real world, there is almost no such thing as 0% or 100%. There are just all the values in between. So, we might find that the evidence in Kiely’s case is really not expected if we assume foul play. It might be low, like 0.05. So in that case, the 0.66 figure gets reduced to 0.033 (0.66 x 0.05), which is very low. But, if the evidence is mostly what we expect to see, we might give it a score of 0.90, and so we will see a figure more like 0.594 (0.66 x 0.90).
Did Foul Play Occur?
Let’s apply BT and try and give a value to the probability that Kiely died from foul play. By foul play, we are allowing for more than just murder. We are allowing for a cover-up of an accidental death: whether that death was due to reckless behaviour (Kiely took unnecessary risks of various kinds); whether it involved the participation of someone else (a ‘friendly’ fight that went on for too long); or whether it was unforeseeable (she slipped on wet ground and fell, hitting her head on a rock).
If Kiely did slip and fall, you would assume that a cover-up would not be necessary. People drown at the beach. Or they get hit by a car due to icy roads or very poor visibility. Tradesmen fall off ladders. Infants die from SIDS. Accidents happen all the time.
However, if an ambulance was called, there might have been a fear that the other party guests would have been questioned and tested for drugs and alcohol. If the party guests and organizers were selfish enough to want to avoid this, then we have a motive to cover-up a death, even though the death was totally unforeseeable, with no moral burden on anyone else.
As we learned from Nick Rinn, the chances of foul play in our scenario are 66%. So, the chances of a car accident are 34%. In this case, when I say “car accident”, that simply means that Kiely drove herself into Prosser Creek Reservoir.
The prior probability that Kiely died of foul play is 0.66. That statistic, as I discussed above, is supplied already. But how expected is the evidence we have if our explanation of foul play is true? The only thing we can do is estimate. We don’t have to be absolutely precise. We can come up with a reasonable number, as long as we are intellectually honest about it.
So let’s try to find a reasonable number. We know it can’t be 0.00. It can’t be a low number, either. Why not? Because we observed the following:
Jagger Westfall, Kiely’s ex-boyfriend, claimed that he was an official filter for information, which PCSO denied.
Sami Smith, a party guest, tried to put herself between other party guests and the police (Teen-to-Teen Talks).
Sami Smith gave an unreliable recollection of the party to Tony Dodge on All American Dream Chaser. This has yet to be completely analysed by a trained statement analyst. Deception Detective, who is a trained statement analyst, did analyse Sami’s interview, but he did not finish the video.
Sami Smith said that she and Kiely were very close friends. But Kate Cuneo, a confirmed friend of Kiely’s, said that Sami wasn’t friends with Kiely.
Sami Smith supplied a night exterior photo of a group of teenagers to the press. This photo turned out to be from a party in Michigan from a previous decade. And some people caught on quite early, before Kiely was found.
Sami Smith claimed that she last spoke with Kiely at 12:36 AM. Yet, the Alder Hill camera seems to show Kiely’s CR-V entering Prosser at 12:33 AM.
We have no photos of Kiely from the Prosser Party.
We have no photos of any kind of the Prosser Party which took place on the night of Friday, August 5th, 2022.
A lot of teenagers who were supposedly at the Prosser Party deleted some of their social media posts. One of the teen volunteer searchers deleted everything in August of 2022 from their Instagram account.
Police supplied a photograph of a security monitor showing Kiely at the Truckee River RV Park on Hirschdale Road, supposedly at around 6 PM on Friday. But they did not supply any of the footage (compare with Brissa Romero, Jolissa Fuentes, Riley Strain, etc.). They did not even tell the public where Kiely was at that time, which could have hampered the search.
We have no solid evidence that Kiely was alive on Friday, August 5th. The implications of this are very concerning.
Police used magnetometers and many other tools and techniques to search Prosser Reservoir. Yet Kiely’s Honda CR-V was not spotted. The car was found in water that was less deep than the length of a BMW 7 Series. At that depth, it would have been visible from the air, given a clear sky.
Adventures With Purpose found Kiely’s car relatively quickly, using fewer tools than police did, suggesting that Kiely’s car was placed in the reservoir after police searches ended.
The autopsy report was sloppy and possibly invalid under State law, indicating the possibility that the body extracted from Prosser Reservoir was not Kiely. Alternatively, the report was forged, and Kiely’s body was never properly examined.
Some people claim that RAW files from the Alder Hill camera were shown to selected investigators. However, even if the cameras produced RAW files (extremely unlikely), nobody else has seen them, which raises suspicion. This is not the same thing as police withholding data to preserve the integrity of the investigation.
We now want a value for P(e|h.b). What score do we give the above? Remember, we’re trying to convert qualities into quantities. How expected is the evidence if our hypothesis (foul play) is true? That’s a lot of evidence that suggests foul play. Not proof, just evidence. Some of the evidence, such as the autopsy report, could be explained as an attempt by officials to make the case go away. In other words, officials want to forget about it, because it’s too hard. But collectively, I think that the evidence very closely accords with foul play. You can give your own score, but I am going with between 70% and 90%.
Now let’s see how well our evidence fits the hypothesis of a car accident with no foul play. There is some evidence we do have that we would expect in this scenario:
Police did not doubt that a party occurred, or that Kiely was there. They claim to have video of Kiely at the party.
The firewatch camera on Alder Hill did capture what seemed to be Kiely’s car entering Prosser Reservoir, at about 12:33 AM on Saturday, August 6th. It captured evidence of human activity in and around The Sanctuary. (Keep in mind that Kiely could have been dead before her car entered the water, but that is what we are trying to solve).
According to the Alder Hill camera, it seemed that one or two cars did drive up to the little peninsula after Kiely’s car supposedly entered the water. This could indicate that some people might have been concerned that an accident happened. Or, they could have checked to see that their plot, to dump Kiely’s body into Prosser, was completed. Again, we want to avoid begging the question. But these possibilities must be mentioned, to acknowledge that the Alder Hill footage doesn’t necessarily show an accident.
We now want a value for for P(e|~h.b). How expected is the evidence if our hypothesis (foul play) isn’t true? There is some evidence that points to a car accident. But not very much. I’m going to give a value between 25% and 33%.
For the calculation, I’m going to use the mean of each set of evidence values. So, for P(e|h.b) I’m using 0.8, and for P(e|~h.b) I’m using 0.29. This is what the equation looks like:
The output, or posterior probability, is 0.528/0.6266, which is about 0.84. Therefore, if our input values are sufficiently precise, there is an 84% chance that Kiely died as the result of foul play.
The lowest output we can obtain, using 0.7 for P(e|h.b) and 0.33 for P(e|~h.b), is 0.8. The highest output we can obtain, using 0.9 for P(e|h.b) and 0.25 for P(e|~h.b), is 0.87.
Afterword
In the first paragraph, I wrote that you could apply Bayes’s Theorem to almost anything, like criminology, pathology, etc. Even literature and art, if you know what you’re doing. But, there are some problems to which you cannot apply it.
For example, what are the chances that God exists, using BT? You would not have the faintest idea how to establish background information for that. Any attempt to establish some kind of basic framework will instantly lead to begging the question, or circular logic, or just pure redundancy (I propose that for the universe to exist, a non-contingent first cause, or God, must be present. And if that is true, then why are we using BT at all?). It’s such a big question that the best we can do objectively is to apply our best reason and logic (ontology).
We can also apply our own personal experiences, but of course we have to be careful. Because some people might not necessarily know the difference between a spiritual experience, a hallucination, or a drug induced psychosis. And this is before we get fancy with ESP, synchronicity, and divine providence. So the God question is either ontological or experiential, but not statistical.
Bayes’s Theorem is a powerful statistical tool with real world applications. It has uses which are classified by the CIA. Some STEM professionals don’t understand it at all. It can even be completely useless in some cases. But, it has legitimate application in criminology and forensic science, where statistics are available and well understood. Kiely’s case is one such application.
It’s my hope that in the future, a statistician or criminologist will take Kiely’s case further, and apply BT more rigorously. Understanding the statistical formula is the easy part. Weighing the evidence, and ascribing quantities to qualities, is the hard part. At the time of publishing, we still don’t have the accident report from California Highway Patrol. That report may very well change our posterior probability in a major way.
Resources and References
‘How We Found Kiely: Murder or Accident?’ by Adventures With Purpose:
(47:09)
Deception Detective’s statement analysis of Sami Smith’s interview with Tony Dodge. DD concluded that Sami was just inserting herself for the sake of attention. This conclusion might be part of the truth, but it is not the whole story. Many people would like DD to not just finish watching the interview, but to also examine this case further:
(1:15:02)
This is the complete interview with Sami Smith by Tony Dodge:
(31:44)
Bayes’s Theorem was used to find a lost sailor, a missing submarine, and millions of dollars worth of gold which was lost before the American Civil War:
(16:18)
Is Bayesian thinking a sham? (Answer: no):
(8:19)
A method to visualize Bayesian probability as a table instead of a formula:
On December 2, 2022, Kas Thinks examined search results for Kiely Rodni’s name. To her surprise, people were searching for Kiely Rodni months before she made the news. How did Kas know this? She put Kiely’s name into Google Trends:
So, is something nefarious going on? The first two things I thought of were:
Psychics had visions or dreams about Kiely, and wanted to see if those were about someone who had already died
Google Trends is corrupt and has serious flaws with its database
Both of those could be true. All of us are psychic, to some degree. ESP, or extra-sensory perception, exists in everyone. Most people don’t have particularly strong ESP, but some do. However, could it be that there are that many people who had visions about Kiely? I am not so sure.
As for Google’s database, it might seem unlikely that a team of experienced programmers don’t know how to index HTML pages or YouTube videos. Remember that YouTube and Google have the same parent company. But, the sum of the codebase that comprises Google is very large and complex. Also, programmers may create perfectly stable code, while at the same time not considering that the choices that they made are the most appropriate ones.
Have you read or watched Jurassic Park? Without spoiling anything, some of the characters found that, although the computer code that automated various functions in Jurassic Park was solid per se, it didn’t take into account anomalous scenarios. In other words, the programmer, Dennis Nedry, made one too many assumptions.
Dennis Nedry wrote all the software for Jurassic Park
It’s worth making a very important point here, which applies to software and to life in general: it’s impossible to predict if your assumptions are the correct ones, because you can’t be presented with every single scenario that could ever happen. Climate models, for example, are notoriously unreliable. But not because the programmers don’t know physics. To start with, their fundamental assumptions may not be appropriate. Secondly, because while they understand known unknowns, they cannot understand unknown unknowns – by definition.
And so, if there is a flaw in Google’s code, we didn’t discover it by looking at the code, but by using it in an anomalous fashion. Were it not for Kiely Rodni, we may not have stumbled upon this problem for a long while – if there is a problem with the code to begin with.
I think that there is a flaw in Google’s code, but I can only solve half of the equation. If people were searching for Kiely Rodni’s name before she died, what information would be returned? Her Instagram account? Her Facebook account? Why not just search social media platforms directly? Isn’t that how you would search for someone?
Also, if potential kidnappers or traffickers knew who she was, why were they searching for her? Surely they already knew where to find her.
Kas has already shown us worldwide results for “Kiely Rodni”. Below is a result just for my country, Australia, for the past 12 months:
You can see that there is a spike that occurs in early 2022. Now let’s try “Kiely Rodni cause of death” just for Australia:
Look at that huge spike in early 2022. It’s bigger than any of the spikes in August. What the hell is going on here? Let’s try that search term, but for the entire world:
That’s more like it, although we still see small spikes around March 2022. But then I looked at search engine results for more clues. And I think I found something. Let’s try Bing first:
See that crappy spam website at the top, next to the date, Feb 15, 2022? Keep that in mind. Now let’s try Duck Duck Go:
Hmm. More spam sites. It looks like they have created search term magnets, where they include any date that you search for, plus your search term. This shouldn’t happen though, surely? Because any website, no matter how crappy, has a creation date. Right? Finally, let’s try Google:
We don’t get those crappy spam sites. So that’s nice. But we get ABC, Fox, and Apple Podcasts. Those sites are used by millions of people every day. And yet those results are being returned with dates before August 6th. But when you click on them, it turns out that those Web pages were not actually created on those dates – 2 August, 5 August, 4 March, etc. It turns out that they just mention those dates somewhere on those pages. And, in one case, ABC writes that she went missing on August 2nd, in an article published in October:
ABC news incorrectly reports that Kiely went missing on August 2nd
But it gets better. I think I found a single clue that can help us understand why our search results are dated earlier than August 6th. Have a look at the last one, dated March 4th. Guess where that links? You might think that it links to the YouTube page of Adventures With Purpose, the group who found Kiely and her car on August 21st. But no, you would be wrong. It links to this video by EWU Crime Storyline:
This video does not mention Kiely Rodni at all (as far as I can tell). The above video has over 5.5M views as of December 7th, 2022. The video by AWP which documents their search for Kiely has 3.4M views. But, look again at the Google result. Do you see what is going on with the search terms? Give yourself a minute.
Notice that the AWP video has 3.4M views. And the link is dated 4 March. Do you see it yet? Google saw the term “3.4” and assumed it was the date, March 4th. For whatever reason, Google doesn’t take into account the creation date of a page, but the text inside the page that might indicate a date. This is not very helpful, because we are not searching for dates, we are searching for terms in a given time bracket. Those are different concepts, and it seems that Google does not differentiate between the two.
As for the problem with Google Trends, perhaps the solution is tied to the above problem, common to Google Search as well as Duck Duck Go and Bing. In other words, what might have happened is this: people searched for “Kiely Rodni” after August 6th, but their searches returned pages with dates that were earlier. Those pages weren’t created before August 6th, they just mentioned dates before August 6th. And Google Trends counts those pages as being created in the past.
So if my hypothesis is true, that Google’s code is flawed, then that explains why it appears that people were searching for Kiely long before she went missing. There is absolutely nefarious activity surrounding Kiely Rodni: the cover-up of her death and disappearance. But there were no traffickers or criminals searching for Kiely Rodni before August 6th.
As for the psychics, I don’t doubt that some people did have visions of Kiely. But from what I understand about ESP, not that many people would have sensed her disappearance. If they did, they might have documented it on a blog somewhere. If you know of such a blog, do let me know.
Compiling the footage of Prosser Reservoir from the Alder Hill fire camera
Heads up Tails up Running to your scallywag
Night falls Morning calls Catch you With my death bag
Prosser Reservoir and its surrounds are beautiful, melancholic and haunting. Most people would not have heard of Prosser were it not for the death of Kiely (pronounced exactly like ‘Kylie’) Rodni, a girl who was a month away from her 17th birthday. From what we know about Kiely, she was happy, vibrant, and musically gifted. Truckee itself, though, is somewhat well known as being one of America’s most haunted towns.
Kiely and her car, a silver 2013 Honda CR-V, went missing after midnight on August 6th. Law enforcement, including local police, CHP and the FBI, supposedly scoured Prosser Reservoir, in the case that Kiely had driven her car into the water. After all, her phone last pinged right near the water. Surely, but sadly, this had to be a typical case of a driver who took a wrong turn and drove straight into the water.
They found nothing, although it came out later that the diving teams were instructed to go no further out than 25 feet. Nonetheless, her car was seen in the water by several people, including Stoney Stone, who used satellite sonar imagery. He found the car on August 7th, and reported that fact. Yet, nothing was done.
CORRECTION: I just discovered that Stoney Stone did not in fact located Kiely’s car. He interpreted an artefact on Google Maps as a sonar image of the car. But in fact this artefact is still there on Google Maps:
Further, this spot is not where Kiely’s car was found.
Nothing, that is, until the Internet called upon Adventures With Purpose (AWP) to investigate. Their claim to fame was locating cars in bodies of water, often related to missing persons cases. Their good work has helped many families find closure when they otherwise would have had none. However, AWP has never before involved themselves with open cases.
AWP was told that Prosser had already been searched, so they searched all the surrounding bodies of water where Kiely’s car might be found. They came up with nothing, and so they decided that they would search Prosser after all. It was their last shot.
They found her car on 21st August, around the middle of the day, 50 feet out from shore and 14 feet down. To give you an idea of how deep the water was, the 2013 CR-V is a little bit longer than 14 feet (slightly more than 4.5m). If you stood the car vertically on the bottom of the reservoir, at the spot where the car was found, the other end would stick up a few inches out of the water.
Not long after that, it was revealed that footage existed of the car going into the water. As it turns out, there are firewatch cameras all over California. The one that captured Kiely’s car going into Prosser is located on Alder Hill. It doesn’t always face Prosser, but thankfully it did on the night of August 5th and the morning of August 6th.
The footage is actually 0.5 fps – one frame taken every two seconds. The images from the firewatch cameras are archived and can be accessed publicly. That is where I got the frames from. You can download the frames yourself from here.
There are a few problems with the footage. One is that it’s a wide shot. Remember that these cameras are firewatch cameras, not surveillance cameras. They exist solely for the public to help in spotting fires before they grow too big. So it’s nobody’s fault that we can only see pinpoints of light at night.
Secondly, and infuriatingly, there is a red light near the site of the Alder Hill camera which turns on and off all night. And it spills into the camera, causing annoying flares. It doesn’t obscure what we need to see but it really is annoying.
Thirdly, the cameras often lose focus when it gets dark. This makes it slightly more difficult to see precisely where light sources are located. It also doesn’t help that the camera is not completely rock steady, and there is a slight about of ‘gate weave’ (even though digital cameras don’t have gate weave per se).
The footage may not seem to reveal much, except to confirm the exact time that Kiely’s car hit the water. However, a reasonable estimate of the speed of the car was calculated by several observers, and that speed was about 10mph. We don’t know if the airbag deployed, but based on the estimated speed, it’s not likely that the impact with the water was enough to trigger the airbag. And in that case, it’s not likely that the car’s Event Data Recorder was triggered (it only records 30 seconds worth of data around an event).
From page 21 of the Honda 2013 CR-V Online Reference Owner’s Manual
We can see a lot of other activity occurring before and after Kiely’s car was submerged. We can see what seems to be people – i.e. car headlights – meeting for non-trivial periods of time. Were they trying to figure out where Kiely was? Did they suspect that she drove her car into the water? If so, everyone knew she was dead not long after half past midnight.
And if people knew what happened to Kiely, nobody said anything to police or to the press. And the search parties, the Teen To Teen talks, all of it, was a smokescreen.
I don’t really want to go into the conspiracy of how Kiely’s friends and family failed her. This is just a place to share my time-lapse videos of the Alder Hill camera. However, make no mistake, even if Kiely simply drove her car into Prosser purely by accident, there is absolutely a conspiracy of silence and deception about it. That is the best case scenario. But I think it’s worse that that.
There are two time-lapses that I have prepared. The first is all the frames from 5th August 4pm to 6th August 6am. The second is all frames from 4th August 3:15pm to 6th August 9am. In both cases, I have overlaid a daylight frame, where appropriate. I have also prepared a blink comparison video, which covers 3.5 minutes up until Kiely’s car’s headlights are no longer visible.
The shorter time-lapse:
The longer time-lapse:
The blink comparison:
I have made my observations, and I will let you make yours. Perhaps in the near future I will make a post about what I think are noteworthy moments.
1961 