NTSB preliminary report here. Location data and weather information included in the report. Weather info was from the airport the aircraft started from (Nut Tree Airport in Vacaville), and generally agrees with data from a nearby MESOWEST station located ~10 miles NorthEast of the lake.
Google Earth View of crash location (North up)
Goggle Earth view of crash location (North up).
The purpose of the flight was an introductory flight for a new employee who was the passenger. The aircraft had not made a water landing before this point in the flight. Obviously, a demo flight would include a water landing. Given the wind direction (30 degrees) and speed (less than 5 knots), the water at the end of the cove had to be as smooth as glass.
My speculation is that the pilot intended to land at the end of cove, misjudged his sink rate over the last piece of shoreline he needed to cross for the landing to the North, revved the engine to gain altitude, but it was too late. The plane hit the shoreline.
The gentle taper of the elevation profile at the end of the cove does not suggest that a rapid climb to get out of the canyon was needed. The observables are consistent with a botched landing attempt.
My guess is that people at ICON can confirm that this location was used for water landings on previous flights to this area.
more May 13, 2017
So a colleague suggested fog formation as a possible contributing factor.
From the NTSB report using the weather at the Nut Tree Airport (origin of subject flight), the air temperature was logged as 18C, and the dew point as 11C. Much too great a difference for fog to form. However, the Nut Tree location is not representative of the conditions at the crash site. (2.5C is actually equivalent to 4.5F.)
Location of Nut Tree below relative to Berryessa (North up).
The Nut Tree is at an elevation of 100' and very close to the San Francisco Bay/Delta. A MESOWEST weather station in Brooks (BKSC1) is slightly closer to the lake, at an elevation of 354' (lake is 500'), and is more representative of conditions at Berryessa.
Location of BKSC1 relative to Berryessa (North up)
A graph of air temperature and wet bulb temperature (dew point) taken from the BKSCI data log is shown below. The time of the accident was approximately 9AM on May 8, 2017. The temperature difference logged below is much more conducive of fog formation than the conditions logged at the Nut Tree Airport. Fog almost certainly formed in the early morning hours, the questions are how long it persisted, if it was present at 9AM, and was it pushed into the area along the West side of Little Portuguese Canyon by the wind?
It is possible that a fog layer formed as illustrated below. The light wind from 30 degrees would compress the fog against the steep bank of the lake. The pilot seeing a clear landing zone descended through what he perceived as a thin fog layer, and believed he was over water at the time since the shoreline was not visible through the fog layer. When he finally got low enough to realize that he had not quite cleared the small shoreline extension into the lake he attempted to ascend, but it was too late. A very reasonable scenario, IMO.
(North up)
Wind speed and gust data from BKSC1 (below) also shows a slightly different picture than the Nut Tree data in the NTSB preliminary report. It was very calm on the morning of the May 8. Even the wind gust speed was below 5 mph which would favor fog formation near the surface of the lake.
Unrelated to the above is a map of Berryessa presumably showing areas where amphibious aircraft can operate (green?). This map would contradict the notion that the pilot was attempting to land in Little Portuguese Canyon. It is not entirely clear what is meant by "seaplane operations". If it means flying, then the pilot was in violation of "permitted only on the main body of water" when he turned up little Portuguese Canyon.
and more May 15, 2017
Close examination of the crash site pictures shows no evidence of scaring in the bank or of a debris "trail" both of which one would expect if there was a significant component of horizontal velocity at the point of impact. This observation would suggest a structural failure or a stall.
However, the picture below suggests that a crash site investigator was examining something South of the crash site (to the left of the crash site in picture below). Possibly a tree impact which would have dramatically slowed horizontal velocity and spun the plane into the bank. Lots of questions remain.
edit May 17, 2017
This is big. From the link below.
"ICON initially suspended all flight operations of the A5 fleet immediately after the accident. Flight operations have now resumed following the NTSB preliminary report."
big
This tells me that a mechanical or structural problem has been largely ruled out by the preliminary investigation, and the data on the flight data recorder. Details of which are not in the public domain as yet.
more May 18, 2017
The picture below has appeared on the Web in a couple of blogs.
Speculation has been made that what appear to be wires in the the 10 o'clock and 1 o'clock positions are a power line that the aircraft ran into and were the cause of the crash.
My own opinion is that these are simply lines used by first responders to secure the wreckage to the shore. Detailed examination of satellite imagery shows no evidence of power lines in this area. BTW, the newest version of Google Earth compatible with the Chrome browser is fabulous.
update May 20, 2017
Well, the Berryessa incident is following a familiar pattern relative to both journalism and the NTSB. Google shows nothing in the last 24 hours, and the ICON corporate website has been turned off. Check back once a month? Icon taking their website down is a big negative for me. I was prepared to regard them in a positive light until then. They are being sleazy now.
edit June 5, 2017
Interesting picture below taken from a location on the opposite shore of the canyon from the crash site. View is to the West. Flight path was from left to right in the photo. Arrow marks the approximate crash site.
more July 17, 2017
No more info from the NTSB (not surprising at this date) or from the manufacturer. The following two screen shots are from the Jeff Wise blog "Aviation Section". Jeff is mainly dedicated to the search for MH370, but did introduce an Icon blog section after the Berryessa incident. The posters below fall into the "take it for what it is worth category", but they sound like they know what they are talking about.
more August 9, 2017
NTSB report news.
Conclusion - pilot error.
NTSB report - final report.
NTSB Report
more September 8, 2017
Lake Berryessa News
