Geebax Posted at 2-6 20:44
I don't think gusts of wind are the cause either, I think it is entirely due to the flight forward movement. One of the seemingly common denominators of this problem is that it usually happens after a period of forward flight,m often in Sport Mode. My theory is that the forward movement causes air to flow into the aircraft, and possibly causing a venturi effect over the aperture in the barometer, thus making the pressure slightly lower inside the barometer chamber.

You mean the cause is not the wind wind? We totally mean the same thing. Gusts get blown in a direction can be simulating the effect of a sudden forward flight (or whatever direction can cause venturi effect inside the mini).

Guorium Posted at 2-6 21:07
You mean the cause is not the wind wind? We totally mean the same thing. Gusts get blown in a direction can be simulating the effect of a sudden forward flight (or whatever direction can cause venturi effect inside the mini).

Yes, that is true, but I suspect the gusts are not the main cause, too many of the cases where this is reported happen after flying in sport mode. On the other hand, I would not rule out gusts.

Geebax Posted at 2-6 21:30
Yes, that is true, but I suspect the gusts are not the main cause, too many of the cases where this is reported happen after flying in sport mode. On the other hand, I would not rule out gusts.

I was saying, gusts and forward movement could both be causing a fast flow of air inside the battery compartment. That is what matters.

Geebax Posted at 2-6 20:40
Just to be clear about this, IMU refers to Inertial Measurement Unit, it is a collection of accelerometers measuring a change in direction in any one of 6 axis, it is not responsible for controlling anything. It is, like the barometer, VPS, GPS and gyros, simply an input to the Flight Controller (FC). Some accelerometer packages also contain the gyros as well.

Which is why I went on to say: "... IMU is supposed to monitor movements inline with XYZ axis and monitor movements around XYZ axis.  IMU should be sending data to drone's flight controller ..."

Geebax Posted at 2-6 20:44
I don't think gusts of wind are the cause either, I think it is entirely due to the flight forward movement. One of the seemingly common denominators of this problem is that it usually happens after a period of forward flight,m often in Sport Mode. My theory is that the forward movement causes air to flow into the aircraft, and possibly causing a venturi effect over the aperture in the barometer, thus making the pressure slightly lower inside the barometer chamber.

Previously I suggested a MM Pilot** needed to test possibility of air flow causing drops.  By flying drone side ways (or backwards) instead of forward.  In attempt to see if problem is related to airflow.  Air being rammed into drone while flying forward, effecting drone's barometer.

Unfortunately, idea of such a test was Trash-n-Trolled, per-usual.

** Pilot with enough experience to handle sudden drop.

Geebax Posted at 2-6 20:44
I don't think gusts of wind are the cause either, I think it is entirely due to the flight forward movement. One of the seemingly common denominators of this problem is that it usually happens after a period of forward flight,m often in Sport Mode. My theory is that the forward movement causes air to flow into the aircraft, and possibly causing a venturi effect over the aperture in the barometer, thus making the pressure slightly lower inside the barometer chamber.

I think you have something there, and as long as I’ve been flying drones it was always known that badly sealed barometer can cause drops and gains in your drone.
I think the member on post #85 suggested tape up all or most of your vents and go flying, which is ridiculous and easy to say when you won’t do it yourself .

Guorium Posted at 2-5 08:34
Nice place! Kinda like where I had my incident. Do you have footage of the 27feet fall?

Nope, however, it is imprinted on my mind.

To add more complexity to this:  I also wonder if the elevation above sea level enters into this.  Perhaps more or less of these problems happen near sea level versus at higher locations.

I was chastised greatly for trying to bring elevation with respect to sea level into another discussion about this a week or so ago.  But the fact is that even though the flight controller is calculating the drone's height with respect to the take off point (applying an offset), the baro unit is always measuring absolute pressure.  And perhaps it struggles at higher elevations.

I just think it may be important to consider the elevation with respect to sea level for each of these incidents in case this has some importance.

The design of the vents and the body of the drone does seem like it could easily change baro readings depending on airflow over the body.  We are dealing with very small barometric changes being interpreted as significant height changes.

Sigmo Posted at 2-10 07:13
To add more complexity to this:  I also wonder if the elevation above sea level enters into this.  Perhaps more or less of these problems happen near sea level versus at higher locations.

I was chastised greatly for trying to bring elevation with respect to sea level into another discussion about this a week or so ago.  But the fact is that even though the flight controller is calculating the drone's height with respect to the take off point (applying an offset), the baro unit is always measuring absolute pressure.  And perhaps it struggles at higher elevations.

If the baro was struggling then all 4 of my high altitude flights around Christmas day should have been ridden with troubles. Among them 2 at over 3500m.  This was the only incident.

Sigmo Posted at 2-10 07:13
To add more complexity to this:  I also wonder if the elevation above sea level enters into this.  Perhaps more or less of these problems happen near sea level versus at higher locations.

I was chastised greatly for trying to bring elevation with respect to sea level into another discussion about this a week or so ago.  But the fact is that even though the flight controller is calculating the drone's height with respect to the take off point (applying an offset), the baro unit is always measuring absolute pressure.  And perhaps it struggles at higher elevations.

I doubt the barometer has trouble with low elevations, they are very precise instruments. I used solid state barometers many years ago to measure the altitude of balloon payloads, and radar verification proved that they were extremely accurate. However, in the smaller scale of things, drone pilots tend to get concerned when the readings are out by 5 to 10 feet, which is minute in the scheme of things and probably outside the capability of barometers. They have been in use in aircraft for decades, although at least the aircraft barometers have an ability to be adjusted for current atmospheric pressure.

Geebax Posted at 2-10 13:44
I doubt the barometer has trouble with low elevations, they are very precise instruments. I used solid state barometers many years ago to measure the altitude of balloon payloads, and radar verification proved that they were extremely accurate. However, in the smaller scale of things, drone pilots tend to get concerned when the readings are out by 5 to 10 feet, which is minute in the scheme of things and probably outside the capability of barometers. They have been in use in aircraft for decades, although at least the aircraft barometers have an ability to be adjusted for current atmospheric pressure.

You mean high elevation? Pretty sure Sigmo implied high elevation where my flight took place.

Guorium Posted at 2-10 19:06
You mean high elevation? Pretty sure Sigmo implied high elevation where my flight took place.

Hmm, I read it differently. Where are you?

This is what happens when people have engineered stuff so well that it channels the creators to tap into the devices easily, command and control.  jk of course..

Trying to find ways on what is going on and finding work arounds is what not to do imo.

Geebax Posted at 2-10 19:51
Hmm, I read it differently. Where are you?

I flew in China. High mountains. 2200-3500m

JJB* Posted at 1-8 00:13
Hiya,

Had a look at your log, see chart.

A second look on my data reflected something very odd. I matched video with the blue ALT trace. The behavior in vidoe matchs the trace. Trouble is, this whole time before the fast drop, I pushed the throttle stick down! Yet the drone kept climbing. By the flags on the bridge and my memory of the walk there the mini was fighting a headwind on its way to the bridge. Probably encountered some translational lift and I tried to hold the altitude down to fight it. I tried to get some low and fast footage at that time. However, why would veritical speed not reflect the change in altitude? I kept gaining altitude but the vertical speed points down between 1240 and 1340. Does v speed include the speed FC thinks that stick down would cause? Because at 1370 when throttle stick returned to neutral the v speed turned positive which matches what the drone was doing.
BTW I think I finally have a clear picture on what is going on now. I think the period of flight between 1240 and 1340 was interpreted by the mini as unwanted ascend (resulted from translational lift). After that window, GPS speed dropped to zero by me cutting the forward pitch. I kept applying down stick til 1370. ALT between 1350 and 1370 the mini did drop as I intended. Then I return the throttle to neutral (kept yaw) but the mini's self-correction of ALT kicked in and it dropped like a rock trying to restore the altitude perhaps at 1230. The control was not lost. I looked closely at v speed just after 1380 and the change of gradient matched the up-stick command from me. So it was reacting immediately but probably not as quickly as I would like it to in the dire situation. Perhaps my lesson here is to be very aware of how the mini's post-flight altitude correction can be amplified to dangerous level by a combination of gusts and high elevation. Leave plenty of buffer zone and be alert.

Guorium Posted at 2-12 05:28
A second look on my data reflected something very odd. I matched video with the blue ALT trace. The behavior in vidoe matchs the trace. Trouble is, this whole time before the fast drop, I pushed the throttle stick down! Yet the drone kept climbing. By the flags on the bridge and my memory of the walk there the mini was fighting a headwind on its way to the bridge. Probably encountered some translational lift and I tried to hold the altitude down to fight it. I tried to get some low and fast footage at that time. However, why would veritical speed not reflect the change in altitude? I kept gaining altitude but the vertical speed points down between 1240 and 1340. Does v speed include the speed FC thinks that stick down would cause? Because at 1370 when throttle stick returned to neutral the v speed turned positive which matches what the drone was doing.
BTW I think I finally have a clear picture on what is going on now. I think the period of flight between 1240 and 1340 was interpreted by the mini as unwanted ascend (resulted from translational lift). After that window, GPS speed dropped to zero by me cutting the forward pitch. I kept applying down stick til 1370. ALT between 1350 and 1370 the mini did drop as I intended. Then I return the throttle to neutral (kept yaw) but the mini's self-correction of ALT kicked in and it dropped like a rock trying to restore the altitude perhaps at 1230. The control was not lost. I looked closely at v speed just after 1380 and the change of gradient matched the up-stick command from me. So it was reacting immediately but probably not as quickly as I would like it to in the dire situation. Perhaps my lesson here is to be very aware of how the mini's post-flight altitude correction can be amplified to dangerous level by a combination of gusts and high elevation. Leave plenty of buffer zone and be alert.

Hi Guorium,

Thinking flights over and use all availabile data (video, eyeball mark1, memory and charts) benefits the awareness that flying  drones is not that easy. And agree with you ; be alert and don`t fly to the max limits of drones and yourself.

About you question 'Does v speed include the speed FC thinks that stick down would cause? "
AFAIK vertical speed does not taken into account RC stick postions, it reflects the actual vertical speed.
Well, it should reflect only the actual vertical speed, otherwise the FC correction calculations can be useless. But if a drone gets out of its normal flight behaviour due to extreem external influences, do hope that all the programmed algorithms can cope with that too....  ;-)

cheers
JJB

JJB* Posted at 2-12 07:34
Hi Guorium,

Thinking flights over and use all availabile data (video, eyeball mark1, memory and charts) benefits the awareness that flying  drones is not that easy. And agree with you ; be alert and don`t fly to the max limits of drones and yourself.

Did not explain why the increase in ALT is accompanied by negative vertical speed. That part of the data is completely bogus without a reason.

Hey Guys,

I'm pretty much new, I was searching before in other threads about the same issue but in my case with the Mavic2 pro, I wasn't able to find anything. This is the only thread where the described problems is just almost the same.    ....... I NEED HELP?

My Mavic 2 pro is losing altitude when I apply forward throttle and this happens in the P mode and sport mode. in tripod mode, it is kinda holding stable altitude. Interestingly it doesn't happen at all at backward throttle whats o ever. it is doing that even so extream that it passes minimum landing altitude almost before contact with the ground so I'm forced to stop and break with the drone. I Have recalibrated all systems several times over the past one month including controller, compass gimbal and IMU over phone and desktop as well. As well do i calibrate my compass before all flights to make sure that no magnetic interferences from the transport might bring my drone in trouble. furthermore, I did try this behavior in different terrains like asphalt and very big grass fields.

I have updated to the latest current firmware version V01.00.0510 and even reinstalled it, but no changes.

I'm pretty much frustrated and don't know, what to do next or else.

djiuser_Kerry7Qtuk5M Posted at 2-14 10:01
Hey Guys,

I'm pretty much new, I was searching before in other threads about the same issue but in my case with the Mavic2 pro, I wasn't able to find anything. This is the only thread where the described problems is just almost the same.    ....... I NEED HELP?

My incident was different from yours. Mine increased altitude first then dropped. I suggest you discuss this in M2P forum and upload your log and video there.

Seems like DJI logics gets spoiled by spurious barometer reading. Fast forward speed (air movement around the drone body) causes a temporary lowered reading that is then processed by the Flight Controller with no or insufficient corretion.

Hi All,

Had a real bad experience the past long weekend.

My son saved forever to buy the new Maverick mini when they came out here in South Africa. It was not cheap for a scholar.
He bean flying for over two years now and do loads of videography for his school. He is definitely not a novice.

So on the weekend he videoed his nephew on the wake board from the boat. Then the next moment the drone started to descent and disappear into the river?
He videoed people behind the boat on numerous occasions and newer before had an issue?
Off-cause he is heartbroken and I am a bit agitated because he lost something so valuable to him and we would not be able to replace it.

He swears its not his fault and desperately want to proof it.
What can I post for the experts so they can make an informed conclusion to what has happened?

Hope someone can help us out.

Tip for Dji: Need to take localisation for rivers, lakes to not land. Like No fly zone. And option for Auto search for land.

Leroux# Posted at 8-14 07:36
Hi All,

Had a real bad experience the past long weekend.

Go to: https://www.phantomhelp.com/LogViewer/Upload/
... where you'll find instructions.
That will give a report you can post a link for the report here.
Or just post the .txt file.

Labroides Posted at 8-14 15:34
Go to: https://www.phantomhelp.com/LogViewer/Upload/
... where you'll find instructions.
That will give a report you can post a link for the report here.

Thank you.
We busy getting the log and will post later today.

Leroux# Posted at 8-14 23:25
Thank you.
We busy getting the log and will post later today.

Hi Leroux,

And mayby start a new thread for your incident / flightlog

cheers
JJB

Yeah well the same thing just happened to me in similar conditions. In my case the mini did hit the water and was completely lost. I was in sport mode, flying over a beach and slowly turning to return home (going above water to do so). The drone was less than 200m away and with a good signal and battery at 70%. When looking at the video, I saw the drone suddenly drop and before I could do anything, signal was lost and there was nothing else to be done. I just submitted a claim under my DJI Care and we'll see what happens.
I usually shoot in cinematic mode and this is one of the very few times I use sport mode. If this is a known issue when flying over water, I would have liked to see a big bright and bold warning...

djiuser_DA0T8LEkshL9 Posted at 8-15 12:34
Yeah well the same thing just happened to me in similar conditions. In my case the mini did hit the water and was completely lost. I was in sport mode, flying over a beach and slowly turning to return home (going above water to do so). The drone was less than 200m away and with a good signal and battery at 70%. When looking at the video, I saw the drone suddenly drop and before I could do anything, signal was lost and there was nothing else to be done. I just submitted a claim under my DJI Care and we'll see what happens.
I usually shoot in cinematic mode and this is one of the very few times I use sport mode. If this is a known issue when flying over water, I would have liked to see a big bright and bold warning...

Sorry for your loss. Do you recall what version of firmware was installed? Also I think the only reason mine did not dip was due to the fact water was shallower than mini's 0.5m landing protection clearance. Mini will stop doing down if it sees solid surface is within 0.5m under VPS. What was depth of water you reckon when your incident happened?

Guorium Posted at 8-15 16:01
Sorry for your loss. Do you recall what version of firmware was installed? Also I think the only reason mine did not dip was due to the fact water was shallower than mini's 0.5m landing protection clearance. Mini will stop doing down if it sees solid surface is within 0.5m under VPS. What was depth of water you reckon when your incident happened?

I upgraded to the latest firmare on Friday so I had the latest version available at the time (whatever that was).

djiuser_DA0T8LEkshL9 Posted at 8-16 06:15
I upgraded to the latest firmare on Friday so I had the latest version available at the time (whatever that was).

Yeah so looks like the latest fw is not an absolute fix. You should uploaded your dat. file from the flight for others to analyse. Don't sync with DJI as that removes the .dat file. After  you extract the .dat file you can then sync your flight so DJI can analyse your flight and maybe give you a replacement

Guorium Posted at 8-16 06:37
Yeah so looks like the latest fw is not an absolute fix. You should uploaded your dat. file from the flight for others to analyse. Don't sync with DJI as that removes the .dat file. After  you extract the .dat file you can then sync your flight so DJI can analyse your flight and maybe give you a replacement

Already synced so no more .dat file for me... Not sure about water depth but it was probably fairly deep. In discussion with DJI customer support. Really hope I will be able to get a replacement.

Colddiver Posted at 8-16 06:42
Already synced so no more .dat file for me... Not sure about water depth but it was probably fairly deep. In discussion with DJI customer support. Really hope I will be able to get a replacement.

In discussion with DJI customer support.
Why not post flight data here to see if anyone can see what it's showing?

Go to: https://www.phantomhelp.com/LogViewer/Upload/
... where you'll find instructions.
That will give a report you can post a link for the report here.
Or just post the .txt file.

Labroides Posted at 8-16 07:23
In discussion with DJI customer support.
Why not post flight data here to see if anyone can see what it's showing?

Managed to upload the flight log. Link is here: https://www.phantomhelp.com/LogViewer/PPH8MIPUK0BYK8AL6XYC

Colddiver Posted at 8-17 08:29
Managed to upload the flight log. Link is here: https://www.phantomhelp.com/LogViewer/PPH8MIPUK0BYK8AL6XYC

Your flight data shows the loss of altitude, but doesn't point to the cause of it.
From 4:53.8 until the end of the flight, about 1:40 later, you didn't make any throttle input to raise or lower the drone.

The Mini always had a poor record for altitude stability and yours was showing a lot of altitude variation in the last 1:40 of your flight as shown in the graph below.
You can see it rising more than 20 ft without any corresponding joystick input with several periods where it came down without any corresponding joystick input.

The final descent from about 20 ft above the water is greater than the altitude variation shown earlier, but that drone was showing a bad case of altitude instability already.

Labroides Posted at 8-17 17:31
Your flight data shows the loss of altitude, but doesn't point to the cause of it.
From 4:53.8 until the end of the flight, about 1:40 later, you didn't make any throttle input to raise or lower the drone.

There was no noticeable altitude variation until the drone crashed. The video was very smooth and level (horizon did not moved). I think that the altitude variations you noticed are caused by the drone going over exposed rock formation (some of them several feet high, we can see some of them on the satellite image) and not because the drone was actually going up or down. It is only after 6m 31.9s, after passing the last rock formation and well into the turn, that the drone was really above water. It is also at that point that the water gets deeper fairly quickly and it looks to me like the drone tried to follow the bottom of the ocean instead of maintaining its current altitude.

Colddiver Posted at 8-18 01:17
There was no noticeable altitude variation until the drone crashed. The video was very smooth and level (horizon did not moved). I think that the altitude variations you noticed are caused by the drone going over exposed rock formation (some of them several feet high, we can see some of them on the satellite image) and not because the drone was actually going up or down. It is only after 6m 31.9s, after passing the last rock formation and well into the turn, that the drone was really above water. It is also at that point that the water gets deeper fairly quickly and it looks to me like the drone tried to follow the bottom of the ocean instead of maintaining its current altitude.

There was no noticeable altitude variation until the drone crashed. ... I think that the altitude variations you noticed are caused by the drone going over exposed rock formation (some of them several feet high, we can see some of them on the satellite image) and not because the drone was actually going up or down.
The data comes from a barometric sensor that measures very small changes in air pressure as the drone ascends and descends.
Flying over rock formations, has no effect on the air pressure.
There is no doubt that your drone was going up and down exactly as shown in that graph.

The video was very smooth and level (horizon did not moved).
That's what the gimbal does for you, the result is stabilised video.

It is only after 6m 31.9s, after passing the last rock formation and well into the turn, that the drone was really above water. It is also at that point that the water gets deeper fairly quickly and it looks to me like the drone tried to follow the bottom of the ocean instead of maintaining its current altitude.
It might look that way to you, but you are just guessing.
Your drone was having trouble maintaining a stable altitude, flying fast in Sport Mode, so it couldn't use the VPS sensors.
It definitely was not following the sea bottom.

This is the same data that DJI will look at.
But they probably won't give as detailed an explanation.
It will be interesting to see what they say.

Labroides Posted at 8-18 03:51
There was no noticeable altitude variation until the drone crashed. ... I think that the altitude variations you noticed are caused by the drone going over exposed rock formation (some of them several feet high, we can see some of them on the satellite image) and not because the drone was actually going up or down.
The data comes from a barometric sensor that measures very small changes in air pressure as the drone ascends and descends.
Flying over rock formations, has no effect on the air pressure.

Thanks for the analysis… Is the original poster’s suggestion that the drone picked up barometric pressure variation it created itself the best theory for what happened here or did I miss something in the manual or flying instructions concerning using sports mode above water?

I will admit being a novice with drones but nothing I recalled doing from the flying tutorials seemed to have  prepared me for handling such a quick drop in altitude. I’ve had this drone for over a year and this was the first time I felt comfortable enough to try the speedier sport mode. Haven’t heard back from support yet but let’s just say I will stay away from sport mode if I can get a replacement…

Colddiver Posted at 8-18 07:06
Thanks for the analysis… Is the original poster’s suggestion that the drone picked up barometric pressure variation it created itself the best theory for what happened here or did I miss something in the manual or flying instructions concerning using sports mode above water?

I will admit being a novice with drones but nothing I recalled doing from the flying tutorials seemed to have  prepared me for handling such a quick drop in altitude. I’ve had this drone for over a year and this was the first time I felt comfortable enough to try the speedier sport mode. Haven’t heard back from support yet but let’s just say I will stay away from sport mode if I can get a replacement…

Is the original poster’s suggestion that the drone picked up barometric pressure variation it created itself the best theory for what happened here
His suggestion was incorrect.
The air pressure around the drone doesn't vary like that.
If it did, everyone's drone would bounce all over the place.

or did I miss something in the manual or flying instructions concerning using sports mode above water?
The issue had nothing to do with flying over water.
The drone was showing significant difficulty maintaining altitude for some time before it got near the water.
And being in SPort Mode, there was no way for the drone to have sensed anything from the water.

I will admit being a novice with drones but nothing I recalled doing from the flying tutorials seemed to have  prepared me for handling such a quick drop in altitude. I’ve had this drone for over a year and this was the first time I felt comfortable enough to try the speedier sport mode. Haven’t heard back from support yet but let’s just say I will stay away from sport mode if I can get a replacement…
It probably wasn't Sport Mode either.
I couldn't see which firmware version you have on your Mini.
Early on there were problems with the Mini not holding altitude.
They were fixed with Firmware Version 1.0.5.
If yours was still on earlier firmware, that could have been a factor.

Labroides Posted at 8-18 07:18
Is the original poster’s suggestion that the drone picked up barometric pressure variation it created itself the best theory for what happened here
His suggestion was incorrect.
The air pressure around the drone doesn't vary like that.

Update on this... Took a while but I got my response from DJI - looks like I qualify for a replacement drone (no details provided from the data analysis)... Not a total loss but I hope I will not experience this again with the replacement (may not ever risk flying at low altitude again)...

I also checked online and it looks like Mavic Mini 2 users also experienced similar issues. Upgrading thus does not seem to be a definitive solution. Does anybody know if more expensive/heavier drones like the Mavic Air are free from this issue? The 2S sure looks nice but investing that much for something that may crash on its own is not very appealing...

Colddiver Posted at 8-25 10:46
Update on this... Took a while but I got my response from DJI - looks like I qualify for a replacement drone (no details provided from the data analysis)... Not a total loss but I hope I will not experience this again with the replacement (may not ever risk flying at low altitude again)...

I also checked online and it looks like Mavic Mini 2 users also experienced similar issues. Upgrading thus does not seem to be a definitive solution. Does anybody know if more expensive/heavier drones like the Mavic Air are free from this issue? The 2S sure looks nice but investing that much for something that may crash on its own is not very appealing...

Don't recall air has such report. It is sad to see mini2 also has reports of altitude issue. Do you have link to those reports?
I think pressure certainly has some thing to do with the issue. I teach students Physics experiments with built-in iphone barometers. Something you would find in drones too. It is very sensitive to height change and should have at least no trouble distinguishing height change around 1 meter which means a height fluctuation of 1m for drone is acceptable (without infrared or ultrasound sensor assisting height measurement).
Drones got other sensors like IMU, algorithm that handle sensor data differently. They can all influence how a drone responses to barometer reading fluctuation. It is without doubt when air flows around mini the pressure will dip. Mini is well known to climb (more vigorously in S mode) while going into a headwind. This post came to the conclusion that the flow of air disturbed the height measurement of mini. Then it attempted to correct height after stopping by descending.
What is new now after I put some more thoughts to it is when I replicated and observed this behavior on grass, it had infrared sensor assisting height measurement all the time so height correction on stop was done accurately. In jiuzhaigou, it was going fast over water so infrared sensor was not working, correction had to rely on a previously determined height (at rest) which was a barometer reading indicating ~0m (from takeoff, roughly 1m above water). So it adjusted to that height giving me the fright. Stick worked the whole time. Not sure why the drop was that fast though. Normally height correction at rest is gentle. Maybe the thin air at 2200m elev did it. Not too sure.
Anyway the lesson here of course is be on high alert when you fly low and fast over water or complex terrain. Preferably you always want VLOS maintained. Especially at high altitude when air is thin and drone needs more time to pick up speed.