Note: In order to provide context for this post on the prosecution’s use of cellphone evidence, Rabia has given me permission to post transcripts from the testimony of Abraham Waranowitz, the prosecution’s expert witness:

• February 8, 2000
• February 9, 2000

The significance of the cellphone evidence that was presented at Adnan’s trial was extremely limited in scope. By the prosecution’s own admissions, the location data from Adnan’s cellphone billing records did not show the cellphone’s location at any particular point on the day of Hae’s murder. Instead, the nominal purpose of the location data was simply to demonstrate that the prosecution’s theory of the case was not disproven by the cellphone records. The prosecution’s expert witness did not prove that the phone actually was or probably was at any location – his testimony was introduced as evidence that it was at least feasible for the prosecution’s case to be accurate.

The prosecution failed in accomplishing even that limited goal, however, for the reasons discussed below. The use of the cellphone data at Adnan’s trial was a jumbled, contradictory mess. The evidence was both used in a misleading fashion, and also factually inaccurate. Whether attributable to confusion or design, the prosecution misreported the expert’s findings in a way that caused the results of his testing to appear consistent with Jay’s story.

It should be noted from the outset that the inaccuracies, flaws, and distortions contained in the prosecution’s cellphone evidence should be attributed to the way that evidence was presented at trial, rather than to the underlying testing performed by the expert, Abraham Waranowitz. Waranowitz’s testing was not the problem. He conducted testing of the performance of AT&T’s wireless network in accordance with the prosecution’s representations (and misrepresentations) about the locations of sites that were relevant to the case, and at trial he testified as to his limited conclusions based on that testing. As Koenig discussed on Serial, “the way the science [was] explained [ ] is right” (Episode 5) (emphasis added). The expert witness presented the science fairly — but the prosecution abused it. Take, for example, this claim made by Prosecutor Kathleen Murphy in closing arguments:

[Waranowitz] also told you that the phone picks the cell tower based on signal strength in the area. He told you, too, that this map shows you — these bright colors each represent areas in which a given tower’s signal strength is strongest. And in these areas, the cell phone is going to talk to the given tower. (2/25/99 Tr. 61-62) (emphasis added).

The prosecutor’s claim was simply false. Waranowitz never said any such thing. Instead, Waranowitz fairly presented the fact that coverage maps are estimates, based on factors that fluctuate and cannot be predicted with any degree of certainty – but in closing arguments, the prosecutor nevertheless presented the cellphone data as conclusive evidence of the phone’s location at the time of a call.

In fact, the cellphone location data was so uncertain and variable that the court almost excluded it from evidence as trial. After listening to Waranowitz’s testimony about all of the variable involved, she asked the prosecution why the cellphone testing should even have been permitted to go to the jury, because given the steep limitations on what it was capable of showing it did not appear to meet the relevancy threshold. In response, the prosecution acknowledged that the cellphone data was severely limited in what it could show, but represented to the court that the only purpose of the test was to prove whether or not it was conceivably possible for Jay’s statements concerning the location of the cellphone to be true:

KU: I will proffer to the Court that when we went to Mr. Waranowitz — when we were talking to him we said, we’ve got cell phone records and we have statements that this AT&T wireless phone were in these locations and these calls were received were made. Is it possible to test the system to see if it is possible for the system to respond in those places in these manners and that was the test to check out the cell phone records and the statements to see if it can be shown that this — that the system can respond in this way. He explained to use you can never say from a cell phone record the spot where something was, you can never prove that. You can only show through the fact that it initiated a call through a cell site, that it was in that coverage area for that cell site. But you can go to specific locations and see if it’s possible for the system to respond as the cell phone records do. That was his test. that was the purpose of it.
THE COURT: Do you believe that the test that you presented to this jury was indicative of what you were trying to prove?
KU: Yes, I do. We asked — we told him what we wanted to show and we left it up to him as to how to do it based on his knowledge of the system. He went out and he did that and this shows that this is — that the system in fact can respond in this manner. (2/09/00 Tr. 17-18.)

Note that, as the prosecution acknowledged, the expert’s findings did not address whether it was plausible that a phone was in a given location at any time, or even whether it was likely to have been in a given location, but rather only whether it was possible for the phone to have been there. The court was not immediately appeased by this reasoning, and asked Urick to explain further why this evidence should even be admitted, given how weak and uncertain it was:

THE COURT: Mr. Urick, why isn’t it irrelevant what a particular phone, an Erickson did on a particular day when the witness has shown through your questions that there are variables that would affect the performance of the phone which effectively makes your test totally irrelevant. If the weather was different, if there were leaves on the trees, if the model of the phone was different. I mean doesn’t in fact everything you’ve presented make the test null and void and irrelevant to this proceeding? You didn’t indicate that you did it with the same — ran the test on a day when the weather conditions were identical to the date that you allege that the body was buried. That the weather conditions were such that there was no ice on the trees, there were no leaves on the trees, that there was no snow on the ground, that the temperature was about the same, that the equipment that was used to test was the same. Your testimony that you’ve presented has not indicated any of the factors that your own witness say affects the performance and thereby doesn’t the test itself become misleading to a jury, to give the impression that the phone that was the subject of this case performed in the same way. I mean granted, the weight of the testimony must be made, a weighing must be made by the jury, but my question is aren’t we bordering on irrelevant versus relevant because the weight is so tenuous here that it might not even weigh anything at all?
KU: The State respectfully disagrees, Your Honor. What we were showing is that yes, it is possible for this system to respond in the way that the cell phone records indicate it in fact did. We are showing the possibility that it did. This is corroborative of other testimony, hence it has great relevance, it goes into corroboration.
THE COURT: So you’re saying no matter how weak or tenuous this might be that it is yet — it is only one item to be weighed in the context with other items that it occurred in the fashion that you’ve presented?
KU: Yes, it is corroborative, therefore it is [ ]relevant.
THE COURT: That is with [Jay’s] testimony that the phone did ring while they were at certain places to corroborate the fact that he was in fact in the areas that you’ve indicated.
KU: Yes.
THE COURT: I see. (2/09/00 Tr. 18-20.)

Ultimately, despite how “weak [and] tenuous” the cellphone evidence was, and despite her remarks that the limitations and unreliability of the data almost made it “null and void and irrelevant to this proceeding,” the trial judge ruled that the evidence was admissible, if of very dubious weight. The judge repeatedly expressed her concerns, however, about the potential of such weak evidence to mislead the jury.

The problems with the prosecution’s use of the cellphone evidence was not limited to erroneous claims made in closing arguments, however. The evidence that the prosecution introduced was itself flawed and inaccurate, as discussed below. However, in order to understand the errors that the prosecution made, we’re going to have to get into the weeds with what exactly the expert witness’s evidence consisted of.

The Expert Witness’s Testing Methodology

Waranowitz was a radio frequency engineer employed by AT&T Wireless Services. As an RF engineer, he was responsible for designing, building, and troubleshooting wireless networks — essentially, his job was to make sure that AT&T’s wireless network was optimally designed, both for purposes of efficiency (not having more towers than necessary for a given location) and call quality (not having a service that pisses off customers). His training and experience in using wireless network test equipment came from training he received from classes provided by AT&T Erickson:

AW: Erickson creates the equipment and the phones that we use, plus training on how to use test equipment that we use to test and optimize the network with and training for design tools. engineering cad, that kind of thing. (2/08/00 Tr. 18.)

The only type of testing experience that Waranowitz had training and experience in performing, as disclosed from his trial testimony, was a form of testing known as drive testing. His testing duties involved determining the best locations for new towers and equipment:

AW: We decided where to put cell sites on what buildings, water tanks and towers. We tested them. we drove them. In other words, what I mean by drive, I mean testing them and we optimized them for performance.
. . .
KU: [D]oes part of your continuing [employment] duties include working out any problems that may arise to the operation of that system?
AW: Yes.
KU: And what do you do in regards to that?
AW: Typically we look at problems that involve interference and dropped calls. This involved us taking out drive test equipment in a vehicle and simulating what the customer sees and identifying the problems and correcting them. (2/08/00 Tr. 20-21.)

Drive testing can be used to test a number of different variables, but a brief summary of what it typically entails can be found here:

For the drive testing performed by Waranowitz at the prosecution’s behest, only a single phone — an Ericcson test phone — was used. His testing of the wireless network consisted of driving around west Baltimore with the prosecutors, making test calls from various locations selected by the prosecution, and verbally stating to the prosecutor what cell towers his equipment had registered a signal from. Here is how Waranowitz’s testing was described at trial:

KU: Now, last fall we asked you to take a test of the system. Can you explain what the test was that we asked you to do?
AW: I was asked to visit a number of places in the area located on this map and take readings and make phone calls to find out what cell site I would originate at certain locations. (2/08/00 Tr. 63.)

During drive testing, cell site information is normally recorded by a computer program that takes readings from the test phone used by the expert, and then correlates the signal data with the phone’s location at the time the signal was registered based on a GPS system. The results of this testing can then be displayed on a map, such as the maps introduced at trial as State’s Exhibits 44 and 45, both of which were created by Waranowitz on the day he performed testing for the prosecution:

(State’s Exhibit No. 44 was marked for identification.)
KU: Have you seen that before?
AW: Yes.
KU: What is that?
AW: This is a — this a drawing, rather a map of the locations I drove and the cell sites that I measured.
KU: And how is that map generated? From what data?
AW: This was generated from an Erickson test phone while driving around in this neighborhood.
KU: Was this generated the day we had you doing the test?
AW: Yes, it was. (2/08/99 Tr. 84-85.)

The map in State’s Exhibit 44 depicts the area around Gilston Park. The road to the far right is I-695, the dashed road running north/south in the middle of the map is N. Rolling Road, and the road running east/west is Route 40.

Waranowitz explained that the numbers displayed on these maps correspond with the tower sector that had the strongest signal at the time the test call was made:

KU: The printed out three digit numbers [on Exhibit 44], do you know what those are?
AW: Those are the frequencies we use.
KU: Explain what you mean by that.
AW:  We have a range of frequencies much like radio and television channels that we use. We record those channels and how strong their signal strength is and relate it to the GPS location. That frequency, it can be identified specifically to a cell site.
KU: And those numbers are printouts of what frequency is being used at any given — at that particular location?
AW: Yes. (2/08/00 Tr. 88-89.)

Based on the maps in Exhibits 44 and 45, which display cell site data that was recorded at short intervals as the testing vehicle was driven to various sites, the type of drive testing performed by Waranowitz involved a test phone that initiated short test calls at frequent intervals, and the tower sector that the call originated on was both displayed to Waranowitz and recorded by the computer.
So to summarize: Waranowitz drove around west Baltimore using an Ericcson test phone that continually initiated outgoing calls and set intervals, and then recorded the strongest signal that the phone received when it made the call, which was identified as a three-digit frequency that corresponded with a specific tower/antenna. The result is a map which shows a series of numbers (each number marking each of the test calls that the phone initiated). For example, in the map above, “973” was the frequency used by L651A, “868” was L698B, and “882” was for L654C. Driving along Rolling Road, south of Route 40, the test phone initiated calls and regular and set intervals, and for each call on that stretch of the road, the test phone recorded 868 (L698B); similarly, north of Route 40 along Rolling Road, test calls recorded five signals from five different towers.

In order to make Exhibits 44 more comprehensible, I have once again used my sweet MS Paint skills to highlight the relevant data:

As you can see, Waranowitz’s testing showed that someone making phone calls in that area could have had calls originating on up to 8 different towers, in a mishmash of overlapping coverage areas.

In all, the prosecution had Waranowitz conduct a drive test of thirteen separate locations. One of the most remarkable aspects of the prosecution’s use of the cellphone evidence at Adnan’s trial, however, was that out of the thirteen locations that were tested, the prosecution had Waranowitz present maps displaying the results of his testing for only two locations. One map was made showing signal strengths around Gilston Park (Exhibit 44, above), and the other showed signal strengths around Cathy’s apartment (Exhibit 45, infra).

For the eleven other locations tested (or rather – for the eleven other locations for which the prosecution disclosed the results of Waranowitz’s testing), the prosecution simply had Waranowitz verbally read out the cell sites that were recorded by his testing equipment as they drove around to various locations, and the prosecutor would then write down what Waranowitz had said:

GC: You didn’t report your findings to them?
AW: I verbally gave Ms. Murphy my readings as we conducted the test. (2/09/00 Tr. 69.)
CG: After you completed all of the tests at the locations that don’t appear on the chart, after you completed them you had reported all of your findings to Ms. Murphy because she was there in the car with you, right?
AW: I reported them as I saw them, yes. (2/09/99 Tr. 159-60.)

By having Waranowitz give verbal reports of his test results – despite the fact the data could be easily recorded and reproduced by his testing equipment – the prosecution thereby limited its discovery obligations to a fraction or what they otherwise would have been.

So why did the prosecution not have Waranowitz also provide maps for the eleven other locations that were tested? Well, one has to assume it is because that data would not have helped the prosecution’s case. Why else would they have chosen to only introduce maps showing cellphone data for regions that have almost no importance to the case whatsoever? Why not a map of Woodlawn? Of Best Buy? Of Leakin Park? Why not show maps of any locations that had anything whatsoever to do with the actual murder and burial? What exactly was the prosecution scared of?

The Results of the Expert Witness’s Testing

Ultimately, the prosecution chose to disclose the following information to the defense in a pre-trial disclosure, which provided that “[i]n an oral statement [dated October 9, 1999], Abe Waranowitz of AT&T Wireless reported the following”:

• Rolling Road at I-70 triggers cell site L651C or L698A;
• [Jenn’s House] triggers L654A or 651B;
• Security Square Mall triggers L651C, although the edges may be L698A (south on Rolling Road);
• [Cathy’s Apartment] triggers L608C or L655A;
• Leakin Park burial site triggers L689B;
• Briarclift Road triggers L648C or 689B;
• Best Buy triggers L651C;
• Crosby at I-695 triggers L654C or L651B;
• I-70 Park and Ride triggers L651B at the west end and L689C at the east end;
• Route 40 at Cook’s Lane up to Forest park triggers L653C on Cook’s Lane; L689C on West Hill, Forrest Park/Park’n’Ride;
• Forrest Park, four blocks east of Security Boulevard, triggers L653C on Cook’s Lane;
• Gilston Park, west of Rolling Road, triggers L698A or B — right underneath; but one gets an L654C reading due to the mound of dirt; and
• Woodlawn High School triggers L651A.

As you can see, for some locations the testing equipment (which was initiating periodic calls) would give more than one result while in a single area. For instance, while driving along the road Jenn lived on, calls were initiated that originated on both L654A and L651B. And, while driving along Crosby road near to I-695, calls were initiated on both L654C or L651B. (In fact, the testing performed at Crosby Road was actually recorded on the edges of the map at Exhibit 44, in the top right corner, which shows where Crosby and I-695 cross.)

Despite the fact that the prosecution only provided the defense with cell data that had been so sanitized and rendered abstract as to be almost meaningless (and then incorrectly informed the defense that no written records of this testing existed), the information contained in the state’s disclosure of Waranowitz’s oral statements was inaccurate and misleading on several points.

Waranowitz’s testing results, as disclosed by the prosecution:

“Gilston Park, west of Rolling Road, triggers L698A or B — right underneath; but one gets an L654C reading due to the mound of dirt.”

Why this was inaccurate:

• The location tested by the prosecution was not Gilston Park. Gilston Park is located to the east of Rolling Road. The prosecution instead tested a vacant lot.
• Even if the prosecution had actually tested Gilston Park, their testing would still have been irrelevant, because Jay never testified that he was at Gilston Park on the day of Hae’s murder. That claim was solely advanced by the prosecution.

Gilston Park is, according to serialpodcast.org, actually what is known as Westview Recreation Area, at the corners of Gilston Park Road and Chesworth Road. But as serialpodcast.org also puts it, “This park stuff is a little strange.” There was no mention of either a Gelston or Gilston Park in Jay’s first interview, in which he claimed that after he dropped Adnan off at track, he went to his house. However, due to the detectives’ “correction” of Jay’s story (as a result of the incorrectly placed L654), Jay gave a different statement, and claimed that after dropping Adnan at track, he first went and smoked a blunt at “Gelston Park” before going to smoke more with Cathy and Jeff at Cathy’s apartment.

Jay never mentioned Gelston Park again in any statement, and every statement he gave after the second interview sticks to the (fabricated) claim that he went to Cathy’s while Adnan is at track. However, once the prosecution realized its L654 error — apparently before the second trial — it changed its theory of the case, and developed a timeline in which Jay, after dropping Adnan at track, went to Gilston Park — not Gelston Park — to smoke a blunt. But this timeline was not based on the testimony of any witness. Despite the fact that Jay testified that he never went to Gilston (or Gelston) Park, the prosecution had Waranowitz perform testing at a location is claimed was Gilston Park, and then had Waranowitz “confirm” that the cell records were consistent with Jay’s testimony.

But the prosecution did not take Waranowitz to Gilston Park for his testing. They screwed it up, and took him to the wrong park all together. As stated in the prosecution’s Amended Disclosure, the testing site was “Gilston Park, west of Rolling Road” — but Gilston Park is east of Rolling Road! Waranowitz’s testimony — and Exhibit 44 — confirm the prosecution’s error:

KU: Did we ask you to go to G[i]lston Park?
AW: Yes.
KU: And can you find that site on this large overlay map?
AW: Yes.
KU: . . . Now, according to that overlay map that site is in — thank you, that location is within the cell site covering from 698, is that correct?
AW: Yes.
KU: What if anything did you discover when you reached that particular location?
AW: I discovered that the phones, 654C were stronger.
KU: And is there a particular geographical location — feature there that caused that?
AW: Yes, there is a small hill there that shadows L698.
KU: And if I may approach the witness I’m going to show you what’s been marked for identification as State’s Exhibit 44. . . . . There’s some circles drawn on there, who drew those?
AW: The circle that I have drawn around G[i]lston Park and parts of Rolling Road indicates that the cellphone that we were using to test saw L654C as its strongest cell site at those locations. . . .  The frequencies that I have circled on this map around Gelston Park and Rolling Road indicate where the phone has picked up L654C as its strongest cell site. (2/08/99 Tr. 84-89.)

Below is another depiction of Exhibit 44, showing Rolling Rock Road (in purple), L698 (marked with a star), and the real Gilston Park (in green):

As you can see, the area circled, in pen, is not even Gilston Park at all. It is an unused field, with no apparent name, and no apparent public use.

But let’s pretend for a moment that the prosecution had actually managed to test the right park. The prosecution’s evidence would still have been misleading. The way the testing of Gilston Park was portrayed to the jury was done in a manner that was aimed at trying to trick the jury into believing that Jay’s testimony matched the cellphone data, by suggesting that Jay’s testimony had been something completely different from what it actually was. Here is how Urick questioned Waranowitz about Gilston Park:

KU: Now, if there was testimony that someone had dropped someone off at school to go to track practice and the person who had the car went to G[i]lston Park, parked for a while and then went back to pick the person up, if you found — and they called at G[i]lston Park, one or more incoming calls were received by the AT&T wireless subscriber telephone and then you found cell phone records that had calls from the L654C cell site, would that functioning of the AT&T network be consistent with the testimony?
AW:  Yes. (2/08/00 Tr. 102.)

Urick knew, however, that there was no testimony whatsoever describing a trip to Gilston Park, and that Jay’s testimony was entirely impossible to reconcile with the cellphone data. Here is what Jay had testified to three days previously:

Jay: We turned and I believe at this time he wanted to get back to track practice because he said he needed to be seen. . . . We were traveling towards Woodlawn High School from Forest Park.
KU: And what if anything happened next?
Jay: I believe we stopped somewhere to buy a blunt .
KU: What is a blunt?
Jay:  I’m sorry. It’s a cigar that you empty and use for smoking marijuana.
KU: And what did you do at that point?
Jay: He didn’t want to smoke with me, so he got out of the car and I left and went to [Cathy’s] house. (2/05/00 Tr. 142-43.)

Cathy’s house. Not Gilston Park. So how could Urick have asked Waranowitz, in good faith, to confirm that Waranowitz’s data was consistent with that did not exist?

I think the most hilarious thing about the prosecution’s cellphone data has to be the fact that, out of 13 separate testing locations, the results from 11 of the tests were apparently so unfavorable and so damaging to the prosecution’s case that it decided it was better off presenting a map of testing that had been performed at the wrong location, at a place where no witness claimed that Adnan’s cellphone had ever been.

But if the prosecution really thought the results from the testing of “Gilston Park” was either the best or second best that had been acquired from Waranowitz’s testing, the results of testing from the other eleven locations must have been really bad indeed.

Waranowitz’s testing results, as disclosed by the prosecution:

“[Cathy’s Apartment] triggers L608C or L655A”

Why this data was inaccurate:

• The prosecution wrongly reported the expert’s testing results in its disclosure of the expert’s oral statements.
• The prosecution’s disclosure provided that the expert witness had made statements indicating that a call from Cathy’s could originate on either L608C or L655A. However, for testing done at Cathy’s apartment, the findings that Waranowitz has reported were also recorded, and incorporated by the computer program that generated Exhibit 45.
• Exhibit 45 confirms that testing performed by Waranowitz showed that test calls made from close to Cathy’s house had originated on both L655B and L608C — but not on L655A.

According to the prosecution’s Amended Disclosure of October 9, 1999, which provided that “[i]n an oral statement this date, Abe Waranowitz of AT&T Wireless reported the following,” Waranowitz’s testing showed that a call made at Cathy’s exact address  could have originated on either L655A or L608C on the day of testing. Which is mighty convenient — because this matches the prosecution’s timeline 6:00 – 6:30 p.m. timeline precisely. During that time period, there were three phone calls that the prosecution claims were received while the cellphone was at Cathy’s house: 6:07, 6:09, and 6:24. One of these calls (6:07) originated on L655A, and the other two (6:09, 6:24) on L608C. (Note: There are two additional calls that Jay claims were received while at Cathy’s house — 4:27 and 4:58. However, neither originates on a tower anywhere near to Cathy’s. The prosecution simply pretended at trial that Jay had claimed these call were received at “Gilston Park.”)

As a result of Waranowitz’s oral statements about his test results (or, at least, as a result of those statements as they were recorded by the prosecutor), the three calls received on Adnan’s cellphone after 6pm on January 13^th appear to support the prosecution’s theory of the case:

KU: Can you identify [State’s Exhibit 45]?
AW: Yes.
KU: And what is that?
AW: This is another drawing or computer generated graphic of the drove that I did on the day of the test.
KU: And the geographical location that that map represents is what?
AW: That would be the address that you gave me.
KU: Now, when you got to the 470 block of Gateway Terrace what if anything did you discover about the functioning of the AT&T Wireless network at that location where you initiated calls?
AW: I found that there were two cell sites very strong in that area.
KU: And affect did that have on your phone?
AW: Depending on where you were located on that road you could originate a call on one of two cell sites.
KU: Now, if you look on 6, lines 14, 15 and 16. There you’ll see three calls. One through cell site 655A, the other two L608C. Were those the cell sites that you picked up in your test when you went to the 2700 block of Gateway Terrace?
AW: Yes. (2/08/99 Tr. 96-97.)

Unfortunately for the prosecution, its disclosure of Waranowitz’s test results was simply inaccurate. We can determine what the actual results of Waraowitz’s testing were because the same data that Waranowitz read out to the prosecutor during the test drive was also recorded by the computer attached to the testing equipment, and that data was used to generate the map introduced at trial as Exhibit 45. That map confirms that, according to the testing that Waranowitz performed, then depending on where you were located on the road where Cathy lived, you could originate a call from either L655B or L608C — but not from L655A, as shown by the frequencies on Waranowitz’s map:

The street Cathy lived on is marked in yellow.

The prosecution’s representations to the court also confirmed, the data that was used to generate Exhibit 45 was, in fact, the exact same data that Waranowitz had orally reported to the prosecutor, and which the prosecutor provided to the defense in its amended disclosure:

THE COURT: The numbers that appear on your State’s Exhibit 44 and 45, numbers like 860, 854, 911, 926, those numbers correspond to what?
KU: We believe it’s a global position, satellite information system that allowed him to place his location when he took a reading and that’s what we believe he will say those are, but these are generated from the global position.
THE COURT: Okay. And the materials —
KM: I’m sorry, Your Honor.
THE COURT: The materials that were used by this witness to make this calculation are they materials or information or notes that were disclosed to the Defense?
KU: If I recall correctly he actually printed this out at the time that he was — we were recording this. So, this was I believe simultaneously I printed up with the oral report that he gave us as to the location. These were the global position satellite was only briefly that day, these were the only two corresponding maps that he was able to print out, but they correspond and I believe were printed out at the same time that he gave us orally what his reading was at each location.
THE COURT: So, the 44 and 45 represents the data that was used to make up your amended State’s disclosure?
KU: Part of it, yes.
THE COURT: When you say part of it, yes, what is the other part?
KU: He had equipment to read the system to tell which cell site the signal was coming from.
THE COURT: No, I understand that, okay. That’s — that’s his internal system?
KU: Yeah. That’s how he got the numbers that he printed out, L6 —
THE COURT: I understand — I understand, but the readings themselves are what appears in 44 and 45?
KU: Yes.
THE COURT: Very well. Your objection is overruled. It’s noted for the record. You may proceed. (2/08/00 Tr. 80-82) (emphasis added).

The conflict between Exhibit 45 and the prosecution’s Amended Disclosure of Waranowitz’s statements can only be explained by (1) the computer inaccurately recording the data reported by the test phone; (2) Waranowitz misspeaking when he was reading out the data in the car; or (3) the prosecutor incorrectly transcribing Waranowitz’s statements. I assume we can reasonably discount the first possibility — but whether Waranowitz read out the wrong number or the prosecution wrote it down wrong, this inaccuracy calls into question all of the test results reported in the prosecution’s amended disclosure.

After all, the testing results were only recorded for two of the sites — and the report of those results was wrong for one of them. It simply cannot be assumed that the prosecution’s disclosure of the results for the other 11 sites was accurately reported, since out of the two test sites for which we can verify the prosecution’s disclosure with the data recorded by the computer, the tower data for one site was inaccurately reported, and the testing at the other site had been performed at the wrong location.

In any event, we can conclude that, if the prosecution’s cellphone evidence has any accuracy at all, then a call received at Cathy’s house could not have originated on L655A, which means that the phone was not at Cathy’s when the 6:07 pm call was received – and Jay was, once again, lying about where the phone was at the time of a call.

Then again, AT&T’s own records plainly provide that “[a]ny incoming calls will NOT be considered reliable information for location.” So who can say where the phone might have been when it received a call at that time?

Waranowitz’s testing results, as disclosed by the prosecution:

“Briarcliff Road triggers L648C or 689B”

Why this data was misleadingly presented:

• The expert found that on the day of testing, a call made on Briarclift Road could have originated on L648C or L689B. This demonstrated that:
  • The cellphone could have been outside of Leakin Park when a call originated on L689B; and,
  • A call can originate on a tower that is several miles from the phone’s location, even when there are five other towers that are closer to the phone.
• The prosecution’s disclosure of the expert’s test results omitted the location(s) along Briarclift Road where testing was performed.
  • This is significant because a call made from that location could have originated on L689B — and the prosecution’s case is almost entirely dependent upon its claim that calls originating on L689B indicated that Adnan’s cellphone was in Leakin Park.
  • By failing to provide the address along Briarclift Road where testing was performed, the prosecution failed to disclose evidence which could have shown how far south beyond the borders of Leakin Park that testing showed a call could have originated on L689B.

Of the 13 locations where the prosecution had asked Waranowitz to perform testing, one of the locations was identified only as “Briarcliff Road” [sic]. Presumably, the road the prosecution was referring to is Briarclift Road — a one-mile long road that stretches from Edmondson Avenue (Route 40) to the southern border of Leakin Park, east of Winans Way. However, the prosecution failed to disclose where along this road that the test was conducted. As a result of his testing, Waranowitz found that, on the day and time the testing was conducted, calls that the test phone initiated while the and prosecutor were driving along Briarclift originated on both L689B and L648C.

Notably, L648 is a tower that is between 2.5 and 3.3 miles from Briarclift — demonstrating that the two-mile estimate of tower range that I have been using on this blog is, in fact, a conservative assumption, and the actual range of these towers can extend beyond the areas shaded in:

In fact, there are five towers in all that are all closer to Briarclift– L651, L653, L654, L652, and L689. Yet Waranowitz’s testing showed that a call from Briarclift could have originated on either L689B or L649C, proving that a call can originate on a tower several miles away from the phone’s location, even when there are five towers that are geographically closer!

In addition to demonstrating that Adnan’s cellphone need not have been close to a tower that a call originated on, Waranowitz’s testing on Briarclift road also conclusively established that a call made from outside of Leakin Park could originate on L689B, the Leakin Park tower. How far outside of Leakin Park does L689B’s range extend? Well, we can’t determine that, because the prosecution’s Amended Disclosure of the expert’s oral statements specified only that testing was done along “Briarcliff Road.” Waranowitz’s testing could have been performed at the border of Leakin Park, or it could have been performed .8 miles away on Edmondson Avenue. The prosecution chose not to inform the defense anything more precise about the test phone’s location.

This omission, in itself, would seem to be significant. One can only imagine what the data would show if the prosecution had chosen to provide maps for the testing done south of Leakin Park, such as the maps that were provided for Cathy’s apartment and (fake) Gilston Park.

But I doubt it is a coincidence that the only two maps the prosecution chose to create for trial were maps of testing done at the two testing locations farthest from L689B. Is this because testing at other sites showed that a call could originate on L689B, even when the phone was far beyond the borders of Leakin Park?

Based on the prosecution’s selective, inaccurate, and misleading disclosures of the testing data, I think that is a fair inference to make.

Waranowitz’s testing results, as disclosed by the prosecution:

“Leakin Park burial site triggers L689B”

Why this data is inaccurate and misleadingly presented:

• The “Leakin Park burial site” was not tested, contrary to the representation made in the state’s disclosure. The expert made a test call while driving down N. Franklintown Road; he did not make a test call from the specific place where Hae was buried.
• The state’s evidence falsely suggests that a call could have been made from the burial site, when in fact the available evidence demonstrates that no reception would have been available at that location.
• Although the expert made test calls from various locations along N. Franklintown Road, no GPS data was provided concerning the locations where these test calls were made.
• The expert’s testimony proved only the following: if a phone was in Leakin Park, then assuming it could get a signal, it would probably originate on L689B. His testimony did not provide evidence to draw any stronger conclusions than that.
• Even if the cellphone could have made a call from the burial site — which, by all available evidence, it could not have — the expert’s testimony concerning the “Leakin Park” calls would still prove conclusively that one of the following must be true: either (1) the prosecution’s cellphone evidence was meaningless junk, or (2) the prosecutions start witness willfully and repeatedly lied under oath in implicating Adnan in Hae’s murder. One of these two things must be true.

Although the prosecution’s case against Adnan would not have existed if it were not for the two calls that originated on L689B — a.k.a., the Leakin Park tower — the prosecution failed to introduce evidence at trial which could have showed that it was actually possible for a cellphone on the AT&T network to have made or received a call from the site where Hae was buried in Leakin Park. Moreover, the evidence that is available demonstrates that AT&T’s wireless network would not have provided coverage at that location, because steep terrain between tower L689 and the burial site would have obstructed the radio signals necessary for a call to have been made or received from there. Because the prosecution failed to acquire evidence necessary to demonstrate that Hae’s burial site would have had reception (i.e., such as by actually making a call from the burial site), there is no evidence that some unexplained topographical feature could, somehow, have provided reception in that area despite the lack of any line-of-site to a tower in the AT&T network.

The cellphone expert’s testimony did not demonstrate that such a phone call would have been feasible, because the expert was not asked to conduct a test from Hae’s burial site itself. The testing was done while in a vehicle along N. Franklintown Road, and there is no evidence that any of the test phone’s periodic test calls had been made when the phone was actually adjacent to the burial site — because the prosecution declined to provide GPS data for testing done in that location, despite the fact it could have easily been made available, as shown by Exhibits 44 and 45. Moreover, as cellphone reception would likely have existed only a few hundred yards to both the east and the west along N. Franklintown Road from where Hae’s body was found, the expert’s data concerning Leakin Park is most likely reporting the signal that was registered by a test call that was initiated either just before or just after the burial site.

For every single testing location for which information was provided about how the testing was done, Waranowitz’s results came from data obtained from a moving vehicle, based on multiple test calls that were initiated at periodic intervals as he drove around with the prosecutor through various streets relevant to the case. Although the prosecution avoided having Waranowitz specify this fact with regard to the Leakin Park test calls, his testimony does confirm he was on the road when the test calls were initiated:

CG: Now, sir, when you went to that place and again, to make sure, you were on the road side of the Jersey wall barriers, correct?
AW: Yes. (2/09/00 Tr. 112.)

In order to be on the “road side of the Jersey wall barrier,” Waranowitz’s testing must have been conducted on the road itself — and based on Waranowitz’s testimony concerning how testing was done at every other location, the car he was conducting the test from was driving down N. Franklintown Road.
Moreover, the prosecution chose not to provide GPS data concerning the location of where the testing was performed in Leakin Park. The prosecution could have done so, easily. In fact, the expert witness did just that for the area around Gilston Park and the area around Cathy’s apartment:

KU: And what system did it use to generate this data from? What’s the global tracking satellite?
. . .
AW: GPS . . . Global positioning system.
. . .
KU: How do you use it?
AW: We use it to specify — we used it to find out where we are when we’re driving around. (2/08/99 Tr. 89.)

Strikingly, however, the prosecution declined to acquire (or, if they did acquire it, to provide) GPS data concerning any testing of the network that took place in Leakin Park. Now, again, why would the prosecution have gone to the trouble of providing the jury with the results of cellphone testing with GPS coordinates for two locations with little relevance, if any, to Hae’s murder, but then declined to provide GPS coordinates for the testing done at the burial site, where the two most important calls to the prosecution’s case were allegedly received? Jay never even testified that Adnan’s cellphone had ever traveled to Gilston Park — let alone made or received any phone calls there. So why did the prosecution have Waranowitz generate a report showing which tower had the strongest signal strength for dozens upon dozens of precise GPS-identified locations in the Gilston Park area?

Well, it is either because the prosecution was oblivious to the significance of the cellphone location data, and erroneously believed that a GPS-backed report of the towers a call in Gilston Park was likely to originate on was more crucial to its case against Adnan than was a similar report for the Leakin Park Area; or else it is because the prosecution believed that a GPS-backed report of the signal strength in Leakin Park would not have supported its case.

Either way, it does not make much difference. The prosecution’s case does not include any data that can be verified through replication, because it failed to give the locations of where testing was performed. Moreover, the prosecution’s disclosure of its testing results contained a serious misrepresentation of fact, and erroneously stated that, contrary to the expert’s testimony, that his testing had been done at the “Leakin Park burial site.” If the defense had assumed that the prosecution had been truthful in its disclosures — and why shouldn’t they have? — they would have been left unaware of the fact that no one had actually tested whether it was possible for a phone at the burial site to make or receive any calls.

Which is a big deal. Because it probably could not have.

I spoke with an RF engineer about the AT&T’s wireless network coverage in Leakin Park, and he confirmed that coverage in that area was “very poor,” and that, moreover, it would not have been possible to determine that a cellphone would have reception in a specific location without actually visiting that specific site and measuring the signal. Short of conducting a test call from the actual location, however, the cellphone coverage for a given location can be examined using Google Earth. By plotting out the coordinates of the tower and the call location, and by examining the elevation profile between those two points, one can then make a determination as to whether it is probable that a phone in that location would have reception, because “[i]f it is possible to communicate with the tower, one should be able to draw an uninterrupted line between the two spots.”

Tower L689 is located on top of the Bernard Mason apartment complex at 2121 Windsor Garden, almost directly north of the burial site. As shown by the tower’s FCC registration, L689 is 94.8 feet (28.9 meters) above ground level. For purposes of the following illustrations, I have rounded the height of L689 to 100 feet above ground level. (Additionally, a web-based topographic profile application is available here, providing a similar review of this data, and providing similar conclusions.)

An examination of the resulting elevation profiles shows that although segments of N. Franklintown Road to the east and west of Hae’s burial site would likely have had reception from L689, the very steep terrain between L689 and burial site itself should have precluded any reception in that area. Note, however, that reception would have been available to the east and west of the burial site:

Elevation profile for burial site. Terrain blocks line of site to tower.

Elevation profile for N. Franklintown Road, east of burial site.

Elevation profile for N. Franklintown Road, west of burial site.

None of this would have been a surprise to the prosecution. They were well aware of the problems with cellphone reception in the Leakin Park area — particularly the areas near Dead Run. Like, say, the burial site.

KU: Mr. Waranowitz, what if any difficulties are there in [Leakin] Park in terms of reception for the cell phone network?
CG: Objection.
THE COURT: Sustained. As any cell phone network or as it relates to AT&T cell phone network?
KU: What if any reception problems as the AT&T wireless network experience in the [Leakin] Park region?
THE COURT: You may answer the question.
AW: The signal strength in [Leakin] Park particularly down where the river and the roads runs through is very weak. (2/08/99 Tr. 125-26.)
CG: That road, however familiar you are with it, runs the breadth of L[eakin] Park throughout Baltimore City, does it not?
AW: I think so, yes.
CG: And, sir, are you aware of the difficulty any cell phone user has regardless of phone, regardless of network of actually speaking on the phone while one is in Lincoln Park?
AW: Yes.
. . .
CG: You are aware, are you not, of the difficulty any cell phone user no matter what cell phone is used in actually . . . receiving a call while one is in L[eakin] Park?
. . .
AW: Yes.
CG: And that’s not a surprise to you, is it?
AW: No, it is not.
CG: Because the terrain in L[eakin Park] Park is difficult, is it not?
AW: Yes, it is.
. . .
CG: The terrain in [L]eakin Park goes to terrain that is much, much higher than the terrain of [N. Franklintown Road], is it not?’
AW: Yes.
CG: And there’s a major stream or waterway that runs through L[eakin] Park, is there not?
AW: Yes.
CG: And you’re aware that the banks of the waterway because the terrain is uneven is often times below what we call street level, referring to the level of [N. Franklintown Road].
AW: Yes. . . .
CG: And notwithstanding that, your network has been launched now for maybe up to three years, L[eakin] Park as a coverage area, it’s always been difficult, has it not?
AW: That is true. (2/09/00 Tr. 114 -116.)

In short, the prosecution was or should have been on alert that it would likely have been impossible for a call to have been placed from the burial site, and yet — or rather, and as a result — they chose not to perform any tests at the burial site itself.

Because if the absence of cell reception at the burial site had been brought to the jury’s attention, it would have been seriously damaging to the state’s case. Jay testified that both Leakin Park calls were received while he and Adnan were digging a hole down by Dead Run. Here is Jay’s testimony concerning the 7:09 call:

KU: When you pulled off and parked the vehicle, what if anything did you take back into the woods with you ?
Jay: Shovels.
KU: About what time of night was this?
Jay: About 7:00 because, like I said, I had paged Jenn and while we were digging, she had called back [at 7:09], and he just told her he was busy now and hung up the phone. We dug for a little bit and he said that’s good enough. We took the shovels
KU: What was the light like?
Jay: It was pretty dark but the moon was out, and I remember there was little bits of snow on the ground. So you could see a little bit. It wasn’t too bad. (2/05/99 Tr. 150-51.)

Oh, Jay. Let’s hold up a second — before we get back to cellphone data, I need to recount just a few of the glaring problems with this testimony: (1) Adnan told Jenn that “he was busy now”? All of Jay’s previous statements had Adnan saying that “Jay is busy now.”; (2) The moon was out? No, no it was not — moonrise was at 4:37 a.m. that night, and was a bare sliver, with around 10% illumination; (3) according to Jay’s story, he has not even paged Jenn at the time he and Adnan were digging the hole — that happens after digging the holes, while Adnan is getting Hae’s body out of the car; and (4) why would Jay and Adnan bring the shovels back to Adnan’s car after digging the hole? They still had to cover up the body! It’s a completely inefficient body burial process.

I digress, though. Back to the cellphone evidence. Here is Jay’s testimony concerning the 7:16 incoming call:

KU: Now, if you will look on the exhibit, lines ten and eleven, there are two incoming calls. While you were back there either digging the hole or burying the body, do you remember him receiving another phone call?
Jay: Yes, but I don’t know who it was .
. . .
KU: Did he speak in English?
Jay: Parts.
. . .
KU: And the other parts, did you know?
Jay: I assume it was Arabic. I mean, I’m not a linguist. (2/05/99 Tr. 152-53.)

Which means according to Jay’s testimony, he and Adnan were at the actual burial site, digging a hole, when the 7:09 and 7:16 incoming calls were received.

No wonder the prosecution was too scared to try and make a test call from the actual burial site. The Leakin Park calls were the prosecution’s entire case – better not let the jury know that Jay was lying about those too.

-Susan