Use of UASs (“Drones”) in 2018 at Kīlauea and Beyond

Video Transcript
Download Video
Right-click and save to download

Detailed Description

This video describes USGS use of Unoccupied Aircraft Systems (UAS) or "drones" for use in 2018 at Kilauea Volcano in Hawaii. The speaker, Angie Diefenbach, is a leading expert in use of UAS for volcano surveillance and research. The presentation was given in June 2020 to colleagues in Ecuador who are interested in applications of UAS at their volcanoes. Angie is a member of the USGS/USAID Volcano Disaster Assistance Program, and many of the developments for use of UAS have come through our international work funded by USAID.
 

Details

Date Taken:

Length: 00:30:23

Location Taken: HI, US

Video Credits

USAID

Transcript

{

\f0\fs24 \cf0 Angie:  I'll be talking about the use of\
\
Angie:    unoccupied aircraft systems or drones during the 2018 Kilauea eruption response,\
\
Angie:    and then I'll end by going into some of the more recent developments using\
\
Angie:    this technology to monitor volcanoes and where we hope to be in the future.\
\
Angie:    It's hard to believe that this eruption was two years ago,\
\
Angie:    and that this time two years ago,\
\
Angie:    I was very exhausted from working on this,\
\
Angie:    but it was a pretty incredible eruption.\
\
Angie:    So just to give you an overview of Kilauea Volcano,\
\
Angie:    I'm sure most everybody is familiar with it.\
\
Angie:    It's a large shield volcano,\
\
Angie:    you have the caldera up here at the summit is what I will refer to Halemaumau lava lake.\
\
Angie:    The middle east rift zone,\
\
Angie:    Puu Oo is right here.\
\
Angie:    Then Leilani Estates or the lower east rift zone is where\
\
Angie:    the focus of lava effusion was in 2018.\
\
Angie:    This is just a close-in view of\
\
Angie:    that lower east rift zone area where lava inundated the residential area.\
\
Angie:    All right. So here's our timeline that's not to scale, but in 2018,\
\
Angie:    Kilauea experienced its largest lower east rift zone eruption\
\
Angie:    in caldera collapse in at least 200 years.\
\
Angie:    After collapse of Puu Oo,\
\
Angie:    then on the 30th of April,\
\
Angie:    magma propagated down rift.\
\
Angie:    By May 3rd, fissure started to open up.\
\
Angie:    There were 24 in total that spanned\
\
Angie:    a stretch of seven kilometers on the lower east rift zone.\
\
Angie:    The following day was a magnitude 6.9 earthquake that was felt up to 500 kilometers away,\
\
Angie:    and it produced five meters of slip.\
\
Angie:    The lava lake began to drain very rapidly at about 2.2 meters per hour,\
\
Angie:    followed by summit explosions that became pretty regularized throughout this episode.\
\
Angie:    On May 11th, Hawaii Volcanoes National Park closed,\
\
Angie:    and around May 16th I got a call from a colleague in Hawaii saying,\
\
Angie:    We know you have a drone.\
\
Angie:    Would you like to come over here and help us because we need to get views."\
\
Angie:    of what's happening at the summit area.\
\
Angie:    We can't see the lava lake anymore,\
\
Angie:    and it seems like it's pretty dangerous for people to be\
\
Angie:    flying over in helicopters because of this explosive activity.\
\
Angie:    So within a couple days, I was able to get a team of five people.\
\
Angie:    At the time, I was the only one in\
\
Angie:    the Volcano Science Center with my drone pilot's license,\
\
Angie:    so I called my friends at the National UAS project office in Denver,\
\
Angie:    Colorado and asked if they would help.\
\
Angie:    They in turn called\
\
Angie:    the overarching aircraft group called the Office of Aviation Services and asked if\
\
Angie:    they would also help.\
\
Angie:    So we got five of us together and we were onsite within\
\
Angie:    two days and flying whenever we could.\
\
Angie:    Lava eruption rate exceeded\
\
Angie:    500 cubic meters per second at its height.\
\
Angie:    When activity focused on fissure 8,\
\
Angie:    it covered an area of 35 square kilometers in total.\
\
Angie:    It was a huge event.\
\
Angie:    It was historically unprecedented in many ways with\
\
Angie:    these explosions and repetitive large-scale collapses at the summit,\
\
Angie:    as well as voluminous lava output in\
\
Angie:    the Lower East Rift Zone that ended around the beginning of August.\
\
Angie:    But it's also unprecedented in that it was the federal government,\
\
Angie:    the US federal government's first response to a volcanic eruption using drone technology.\
\
Angie:    Even though this technology has been actively used\
\
Angie:    by geoscience communities for probably over a decade now,\
\
Angie:    it really hadn't taken hold here with the US Geological Survey,\
\
Angie:    except with VDAP and our international partners and we've been\
\
Angie:    using that technology pretty extensively for over five years now.\
\
Angie:    It's important to note that the eruption took place with\
\
Angie:    two very distinct styles in two different locations.\
\
Angie:    So it made it pretty hard to be at both places at once.\
\
Angie:    Pretty much impossible, especially with a small team\
\
Angie:    of five pilots where we had to stay together.\
\
Angie:    We had the explosive activity or the caldera collapses\
\
Angie:    happening at the summit and then the voluminous lava output in the lower East Rift Zone.\
\
Angie:    We used various drone platforms and sensors for this response.\
\
Angie:    Our biggest one was the Firefly Pro 6,\
\
Angie:    which is a fixed wing vertical takeoff and landing platform\
\
Angie:    that we equipped with both thermal and IO cameras for mapping purposes,\
\
Angie:    pretty much extensively used at the summit.\
\
Angie:    We also had little small quad-copters,\
\
Angie:    the Mavic pro and the Autel EVO,\
\
Angie:    which were used predominantly for\
\
Angie:    situational awareness and I'll go into that a little bit later.\
\
Angie:    Then the Matrice 600,\
\
Angie:    the one on the bottom here with the hexacopter was the workhorse of the eruption.\
\
Angie:    We used it for lots of different things including mapping.\
\
Angie:    We attached two different gas sensors to it, thermal cameras.\
\
Angie:    So we were able to do concurrent monitoring with that one.\
\
Angie:    So the goal of these UAS surveys or drone surveys were to provide a stream of\
\
Angie:    rapid turnaround data products to scientists and\
\
Angie:    emergency managers to make decisions for public safety.\
\
Angie:    So I'm going to go into how these were used throughout the eruption\
\
Angie:    for various types of monitoring.\
\
Angie:    So we use them to look at gas emission rates, composition, and concentrations.\
\
Angie:    We would fly this right into the plume of Fissure 8,\
\
Angie:    surprisingly didn't lose any of our drones because this is a very volatile environment,\
\
Angie:    over saturated the equipment that we had.\
\
Angie:    It was the only way they were able to get gas measurements from the plume,\
\
Angie:    it would've been way too dangerous in manned aircraft.\
\
Angie:    You can see the river of lava below.\
\
Angie:    There's a video here. Moving a little fast. There we go.\
\
Angie:    So we were able to integrate two USGS developed gas sensors.\
\
Angie:    We have on the top right here,\
\
Angie:    the micro DOAS and on the bottom multi-gas.\
\
Angie:    We use them both at the summit,\
\
Angie:    obviously as well as Lower East Rift Zone.\
\
Angie:    Lower East Rift Zone is where we were able to get  the best measurements.\
\
Angie:    We have this big hefty plume that was there throughout.\
\
Angie:    As soon as fissure eight became the main source of lava.\
\
Angie:    The micro DOAS measurements were really difficult to get with\
\
Angie:    a drone because the plumes were so massive.\
\
Angie:    So, you had to go from clear air under the plume to clear on the other side,\
\
Angie:    and at one time it was about five miles or several kilometers wide.\
\
Angie:    So it was really hard to get those measurements,\
\
Angie:    but the multiGAS turned out to be very efficient.\
\
Angie:    We also used the drones to capture\
\
Angie:    very stable nadir videos to support eruption rate measurements.\
\
Angie:    The nice thing about their DJI platforms is that they just hover perfectly still,\
\
Angie:    so we got this beautiful video.\
\
Angie:    We also had the exact GPS coordinates of where they would go,\
\
Angie:    what we call benchmarks,\
\
Angie:    and we could reoccupy those at any given time\
\
Angie:    along the channel, that went out to the ocean.\
\
Angie:    So here's just a graphic from\
\
Angie:    Hannah Dietterich's paper that's in prep right now where she used\
\
Angie:    UAS video as well as DEMs that we created to look at eruption rate measurements.\
\
Angie:    Like I said before,\
\
Angie:    at the height during pulsing activity of fissure eight,\
\
Angie:    we're measuring up to 500 cubic meters per second coming out of fissure eight.\
\
Angie:    It was pretty wild. It was pretty steady at\
\
Angie:    about 150 cubic meters per second throughout the main part of the eruption.\
\
Angie:    Although our team was there to really support scientific monitoring that was going on,\
\
Angie:    it was quickly realized that we could also help\
\
Angie:    with public safety and emergent public safety situations.\
\
Angie:    One way that we did this is using technology that was provided by NASA,\
\
Angie:    we are able to live stream videos to\
\
Angie:    emergency operation centers in Hilo and Honolulu.\
\
Angie:    Actually that video could be looked by anyone in the world at any given time.\
\
Angie:    Really, our emergency managers had on-demand 24/7 access\
\
Angie:    to real time situational awareness video so they can make quick decisions on.\
\
Angie:    One example was on the night of May 27th.\
\
Angie:    We had been out there mapping the lava flow extent all day and realized that\
\
Angie:    magma was pooling up between two\
\
Angie:    of the fissures and it looked like it was about to spill over.\
\
Angie:    We said, okay, we have one set of batteries left, let's go check this out.\
\
Angie:    Sure enough, lava began to spill over.\
\
Angie:    It was rapidly going into an occupied residential area.\
\
Angie:    We were able to provide this real-time video feed to first responders to get people out.\
\
Angie:    We actually were able to use a drone to help get a resident to safety.\
\
Angie:    He was stuck and disoriented,\
\
Angie:    didn't know how to get out.\
\
Angie:    First responders were having a hard time finding his location.\
\
Angie:    We flew the drone overhead,\
\
Angie:    and said follow it to the first responders so\
\
Angie:    he could see the light of the drone overhead,\
\
Angie:    and we were able to get in there before his house became engulfed in lava.\
\
Angie:    We were also able to provide critical information\
\
Angie:    on critical infrastructure and related impacts.\
\
Angie:    On the right here,\
\
Angie:    we can see one of the cell towers that had been knocked down by lava flow.\
\
Angie:    Various government agencies requested us to go in and look at property damage assessment,\
\
Angie:    so it wouldn't take months,\
\
Angie:    and months for people to get insurance on their damaged properties.\
\
Angie:    We provided county officials with daily updates on lava flow boundaries as well.\
\
Angie:    One key thing is that we were using the drones to also\
\
Angie:    monitor topographic change throughout the eruption,\
\
Angie:    both at the lower east rift zone and the summit.\
\
Angie:    We were able to utilize structure from motion and\
\
Angie:    photogrammetry workflows in the field on a laptop,\
\
Angie:    and we provided a stream of\
\
Angie:    rapid data products to people who needed to make quick decisions.\
\
Angie:    In the lower east rift zone,\
\
Angie:    when we would fly over an area,\
\
Angie:    we had a goal of returning a DEM,\
\
Angie:    an orthomosaic, and a point cloud to\
\
Angie:    emergency managers within 45 minutes of finishing that flight.\
\
Angie:    At the summit, we had a turn around of about 24 hours just because it was\
\
Angie:    such a massive area to cover and processing was difficult.\
\
Angie:    Images were geotagged with RTK GPS information that was on the platforms,\
\
Angie:    and we used flight planning software to go and re-occupy certain sides.\
\
Angie:    At the summit, to create a DEM of that total area,\
\
Angie:    we used the fixed-wing drone and it took about seven flights,\
\
Angie:    and each flight was 30-40 minutes,\
\
Angie:    so it took a lot of flying to get that coverage.\
\
Angie:    So on the lower east rift zone,\
\
Angie:    these DEMs were used to help map out\
\
Angie:    the lava flow advancement boundaries that you see in the left graphic through time,\
\
Angie:    and the updated topography in the lower east rift zone\
\
Angie:    was also used rapidly and integrated into\
\
Angie:    lava-flow-path-likelihood  software that\
\
Angie:    helped emergency managers make evacuation decisions rapidly.\
\
Angie:    So that's on the right, we have updated a drone topography\
\
Angie:    on top of previously acquired LiDAR imagery that helped\
\
Angie:    people figure out where inundation may occur next and whether we needed\
\
Angie:    to shut down important facilities like a geothermal plant that was still in operation.\
\
Angie:    After the eruption was over,\
\
Angie:    I was able to take a weeks worth of different drone flights to\
\
Angie:    produce an almost complete topographic map of lava inundation.\
\
Angie:    You can see here in the purple,\
\
Angie:    are areas that we're missing.\
\
Angie:    So we could only fly these for about 30 minutes at\
\
Angie:    a time and cover relatively small patches and then we'd have to mosaic them together.\
\
Angie:    As the eruption advanced,\
\
Angie:    we lost lots of different launch sites that we had been flying from.\
\
Angie:    So by the middle of the eruption,\
\
Angie:    we had three spots that we could fly from.\
\
Angie:    One was in Leilani Estates,\
\
Angie:    one was here at this Y,\
\
Angie:    and one as over here at this cinder cone.\
\
Angie:    So following the eruption, putting those together,\
\
Angie:    we get a total erupted volume of around 600 million cubic meters on\
\
Angie:    land and an additional 600 million cubic meters or more went out into the ocean.\
\
Angie:    You can see that's where the thickest part is here.\
\
Angie:    So in total it's about 1.2 cubic kilometers of lava was erupted,\
\
Angie:    and that's a minimum number.\
\
Angie:    The mean output rate was about 150 cubic meters per second,\
\
Angie:    and from Fissure 8, that average was around 175 cubic meters per second.\
\
Angie:    But again, during surging events at Fissure 8,\
\
Angie:    we exceeded 500 cubic meters per second of lava output.\
\
Angie:    We also performed topographic monitoring of the caldera collapse.\
\
Angie:    I'm going to show you here a time lapse footage that's\
\
Angie:    from the Hawaiian Volcano Observatory looking out over the caldera,\
\
Angie:    and some things you note is just looking at this back wall will drop down.\
\
Angie:    Subsidence was episodic.\
\
Angie:    Type A events occurred every 24-48 hours and at\
\
Angie:    times the slumping block moved more than eight meters during each event.\
\
Angie:    So let me play this video for you.\
\
Angie:    Let's see if this works. Here we go.\
\
Angie:    So this is from April to August,\
\
Angie:    and it shows the entire extent of caldera collapse.\
\
Angie:    We see the summit explosions,\
\
Angie:    and then there, it starts to collapse.\
\
Angie:    We were able to create a topographic time series of caldera growth,\
\
Angie:    and you can see here is the volume change through time.\
\
Angie:    We were lucky on one occasion to be able to\
\
Angie:    produce a DEM before and after a type A event.\
\
Angie:    On average, about 13 million cubic meters of subsidence occurred between each event.\
\
Angie:    The total volume collapse was just under a cubic kilometer at the summit,\
\
Angie:    averaging about 144 cubic meters per second.\
\
Angie:    I'll just show this. Here's a time lapse of our shaded relief images from those DEMs,\
\
Angie:    and some of them are partial.\
\
Angie:    We had to contend with eruption columns and bad weather a lot of times.\
\
Angie:    But overall, it's a pretty great dataset.\
\
Angie:    The caldera ended up subsiding more than 500 meters.\
\
Angie:    Up here on the left is\
\
Angie:    a shaded relief image from a 2009 LiDAR survey of the summit.\
\
Angie:    This little red circle indicates where the lava lake was.\
\
Angie:    Here is from August 2018 after the collapse events have topped.\
\
Angie:    This is where the lava lake would be at present day.\
\
Angie:    You can just see massive deformation has occurred.\
\
Angie:    Here's just some more images that highlight the deformation field.\
\
Angie:    On the left here are obliques of orthophoto mosaics that are draped\
\
Angie:    on to DEMs before the caldera collapse started.\
\
Angie:    Here's the parking lot where you would park to go to the Halema'uma'u Overlook.\
\
Angie:    There was a lava lake there.\
\
Angie:    That parking lot now sits way down on the drop block,\
\
Angie:    and the road has been highly deformed.\
\
Angie:    Then on the right here,\
\
Angie:    is looking at the different deformation fields from\
\
Angie:    full scale where you have over 500 meters of subsidence.\
\
Angie:    To high levels of deformation, looking at intermediate,\
\
Angie:    you can see the incipient slumping here,\
\
Angie:    and then to low levels of deformation where you can actually\
\
Angie:    see changes in the rate of deformation on the caldera fault.\
\
Angie:    You can also identify areas that are relatively stable,\
\
Angie:    so it's easy for us to measure between successive DEMs very accurately.\
\
Angie:    We've had issues with alignment just because we were not able to\
\
Angie:    put out ground control because it was an unsafe environment.\
\
Angie:    Putting a fine transformation to these DEMs to get them to align much better.\
\
Angie:    So you can see this is HVO in the shaded relief image\
\
Angie:    and these are offsets we would see between successive data sets,\
\
Angie:    and then we've been able to align them now since the eruption's\
\
Angie:    been over to get high quality deformation measurements.\
\
Angie:    This is just showing you the results.\
\
Angie:    You can see some noise here,\
\
Angie:    and that's because we are looking at LiDAR versus photogrammetry derived DEMs.\
\
Angie:    The vegetation kind of throws your eye off.\
\
Angie:    But if you look at a facility here like HVO, you will notice that it barely moves.\
\
Angie:    Here just again, showing before the summit of Kilauea,\
\
Angie:    you'll notice HVO is up here and then what it looks like present day.\
\
Angie:    Topographic monitoring was really critical during the eruption.\
\
Angie:    We produced or flew about 300 mapping missions,\
\
Angie:    produced an assortment of topographic models.\
\
Angie:    Again, the total erupted volume that we get from\
\
Angie:    both photogrammetry and bathymetry data is about 1.2 cubic kilometers of lava,\
\
Angie:    and the caldera collapses is just over 800 or just under a cubic kilometer.\
\
Angie:    Again, those mean output rates are actually quite similar.\
\
Angie:    The drone technology was really instrumental in\
\
Angie:    monitoring deformation throughout the eruption,\
\
Angie:    both at the lower East Rift zone and the summit,\
\
Angie:    as well as providing critical gas measurements,\
\
Angie:    situational awareness, and lava advancement rates, as well as boundaries.\
\
Angie:    We have the densest time series of caldera collapse and a fissure eruption ever created,\
\
Angie:    thanks to this technology.\
\
Angie:    One of the big highlights I think of using drones for\
\
Angie:    monitoring volcanoes is that we have operational control of an aircraft,\
\
Angie:    allowing us to survey areas on demand.\
\
Angie:    Whereas when we fly with other types of manned aircraft,\
\
Angie:    helicopters, airplanes, we have to wait for them to be ready to go out.\
\
Angie:    Whereas a drone, we can fly 24 hours a day,\
\
Angie:    daytime, nighttime, whenever we need it.\
\
Angie:    There were of course, lots of limitations and restrictions to this response.\
\
Angie:    One on the scale of the eruption was quite problematic.\
\
Angie:    As you recall,\
\
Angie:    the eruption was geographically separated from the summit to the lower East Rift zone,\
\
Angie:    so it was difficult to travel to both places each day.\
\
Angie:    The platforms we were using, the drone technology,\
\
Angie:    has limitations in range or how far they can fly,\
\
Angie:    as well as how long they can be in the air,\
\
Angie:    and we had rapidly changing topography.\
\
Angie:    Even though we would survey areas and have a 45 minute turnaround time for DEMs,\
\
Angie:    by the time we sent it off to emergency managers,\
\
Angie:    it was already outdated.\
\
Angie:    The launch points at both eruption sites,\
\
Angie:    one was consumed by lava and the lower East Rift zone,\
\
Angie:    so we were limited to three places to fly from,\
\
Angie:    and at the summit we lost some of our launch sites as well\
\
Angie:    because the roads would just crack open and we can no longer get to them.\
\
Angie:    Flight conditions were often very difficult with eruptive activity or fog, wind, rain.\
\
Angie:    Like I said before, ground control was an issue because it was really too\
\
Angie:    dangerous to get out there and put ground control targets on the ground.\
\
Angie:    Even though we had RTK GPS technology on our platforms,\
\
Angie:    we still had errors in alignment.\
\
Angie:    Pilot resources as well.\
\
Angie:    I mean, we were trying to fly 24 hours a day.\
\
Angie:    After about a month of this response,\
\
Angie:    we switched to having two different teams.\
\
Angie:    One nighttime team, one daytime team,\
\
Angie:    but still when you have five people,\
\
Angie:    you have to have somebody fly,\
\
Angie:    somebody process the data.\
\
Angie:    Then data delivery and data management was a huge issue.\
\
Angie:    Trying to get data quickly into the hands of people who need\
\
Angie:    it, but being held back by things\
\
Angie:    like firewalls and access to Department of Interior Google drives, things like that.\
\
Angie:    It was difficult. It started out as a five-person rapid response team.\
\
Angie:    Soon expanded to include 36 drone pilots from across the Department of Interior,\
\
Angie:    including the bureau's USGS, BLM, OAS.\
\
Angie:    Our response lasted four months.\
\
Angie:    It actually went one month beyond when the eruption ceased,\
\
Angie:    just because of the potential for reactivation.\
\
Angie:    We flew over 1200 flights with more than 300 hours of flight time,\
\
Angie:    and collected about six terabytes of data.\
\
Angie:    Since the eruption's been over,\
\
Angie:    we've been able to reprocess the data at\
\
Angie:    much higher quality and that number has probably doubled.\
\
Angie:    It's the first federal UAS response to a volcanic eruption in the US,\
\
Angie:    and one of the largest,\
\
Angie:    if not the largest drone response to a natural disaster in the United States.\
\
Angie:    It truly showed the utility and value of this type of technology as\
\
Angie:    an additional observation and sensing\
\
Angie:    platform for volcanic studies and eruption responses.\
\
Angie:    Like I said before, the technology had not really\
\
Angie:    taken off with our domestic program very much,\
\
Angie:    but this response really showed that it is worth having in our toolkit.\
\
Angie:    We already knew it was valuable from our work\
\
Angie:    with colleagues around the world and VDAP,\
\
Angie:    but it took an eruption like this to really open the eyes of some of the managers.\
\
Angie:    Moving forward, because people saw that this was a great tool to have.\
\
Angie:    We now have a UAS or drone project within our Volcano Science Center.\
\
Angie:    We have a team-wide resource of pilots, aircraft and sensors.\
\
Angie:    The Hawaiian Volcano Observatory is now operational with\
\
Angie:    their drone technology and being able to fly it.\
\
Angie:    Last October, we were able to train\
\
Angie:    seven additional pilots from across the Volcano Science Center.\
\
Angie:    We've learned a lot from this,\
\
Angie:    so hopefully we'll be better prepared for the next eruption response,\
\
Angie:    and our federal government is looking into having\
\
Angie:    a drone response plan for natural hazards across different mission areas.\
\
Angie:    So now beyond that eruption,\
\
Angie:    most of you may know that the lava lake\
\
Angie:    disappeared but a water lake appeared at the summit of Kilauea.\
\
Angie:    Last July, water was identified in the Halema'uma'u crater.\
\
Angie:    It's been rising about one meter per week since it was first identified,\
\
Angie:    and SO2 emissions measured by both car traverses as well as drones were really low,\
\
Angie:    and it was often wondered whether or\
\
Angie:    not the gases were just being scrubbed by this lake that had appeared.\
\
Angie:    Kilauea has a history of explosive activity and\
\
Angie:    this really highlights that there could be a phreatic eruption potential.\
\
Angie:    Because we didn't know if gases were pooling down in this deep crater that's over\
\
Angie:    500 meters deep and there's rock slide activity,\
\
Angie:    it's still a pretty dangerous environment,\
\
Angie:    drone technology was the obvious tool to go down and get a sample of this lake.\
\
Angie:    Right now, the lake is approximately 30 meters deep,\
\
Angie:    and this is when it was first identified as this tiny little green dot here.\
\
Angie:    This is what it looked like last month.\
\
Angie:    It changes color.\
\
Angie:    It's pretty hot and we wanted to get a sample,\
\
Angie:    so we are going to use this new technology to do that,\
\
Angie:    using a hydra sleeve payload.\
\
Angie:    Here's what it looks like.\
\
Angie:    We had about a 30 foot rope attached to the hexacopter,\
\
Angie:    the Matrice-600, and it was flagged at five feet intervals.\
\
Angie:    Then we have this plastic sleeve with a funnel at the top\
\
Angie:    and a temperature probe attached to the bottom.\
\
Angie:    That actually added weight so it would sink down into the water.\
\
Angie:    You get about 750 milliliters of the sample for each sleeve that we used.\
\
Angie:    For perspective, this is a zoom shot of that drone getting closer to the lake.\
\
Angie:    It's just a tiny speck,\
\
Angie:    it's really hard to see\
\
Angie:    and basically impossible to see without binoculars\
\
Angie:    once you get down there to take a sample.\
\
Angie:    Here is a video of one of our missions\
\
Angie:    and I'm just going to fast forward just a little bit, I think.\
\
Angie:    There's the hydrasleeve on the long rope.\
\
Angie:    If anyone's interested in seeing these videos in their full duration,\
\
Angie:    I'm happy to share.\
\
Angie:    But here, you can see a clear line in the water.\
\
Angie:    You have where the groundwater is coming in to the lake versus over here where it's not,\
\
Angie:    green color and orange color here.\
\
Angie:    We took a pretty slow descent to make it safe.\
\
Angie:    But here, we're getting closer.\
\
Angie:    My apologies for not editing this earlier,\
\
Angie:    but I was just able to get it downloaded before this talk, and let's see.\
\
Angie:    Still not going in. Anyway, it dunks down,\
\
Angie:    captures a full sample, and then you have to be very careful in\
\
Angie:    flying it back because of the pendulum effect on such a long line.\
\
Angie:    You have to fly very slow and steady in getting that back.\
\
Angie:    There it is, submerged into the water, and again,\
\
Angie:    that flagging helped us identify the depth that we were sampling.\
\
Angie:    What we found out,\
\
Angie:    the pH is relatively neutral at 4,\
\
Angie:    but after our chemical analyses,\
\
Angie:    the sulfate concentrations are seven times higher than\
\
Angie:    the nearby well that's about a kilometer away,\
\
Angie:    which potentially says that there\
\
Angie:    could be significant amounts of SO2 dissolving in the lake,\
\
Angie:    since we're not getting very much from our airborne surveys.\
\
Angie:    The temperature is about 70 degrees Celsius.\
\
Angie:    We've been able to do two of these sampling missions.\
\
Angie:    One in October of last year and one in January of this year.\
\
Angie:    You can see the HVO set up\
\
Angie:    a temporary little geochemistry lab just off of the edge here of the caldera,\
\
Angie:    and this is kind of the scale,\
\
Angie:    and then here's how we would capture the sampling load before takeoff and after.\
\
Angie:    What we had hoped is that we'd have\
\
Angie:    routine US sampling to monitor the evolution of the lake.\
\
Angie:    But following our training session in October of last year,\
\
Angie:    the Department of Interior has grounded all drone operations.\
\
Angie:    We've been able to get emergency exceptions to fly,\
\
Angie:    but it's not like we can go out and do this too routinely,\
\
Angie:    and the pilots at HVO are just newly minted drone pilots,\
\
Angie:    and this is quite a difficult mission.\
\
Angie:    We've had some setbacks in addition to the COVID-19 restrictions.\
\
Angie:    Our last sample was in January.\
\
Angie:    We're hoping to get out again this summer to capture another one.\
\
Angie:    Additional developments on the UAS front,\
\
Angie:    we're working with lots of different technicians within\
\
Angie:    the USGS on improving payloads and sensor developments,\
\
Angie:    especially with respect to the water sampling mechanism,\
\
Angie:    how we drop it, how we have a temperature probe on it.\
\
Angie:    One mission we'd like to do,\
\
Angie:    at Kilauea,\
\
Angie:    is to characterize  the newly exposed caldera walls,\
\
Angie:    to better constrain eruptive history by doing some high resolution\
\
Angie:    imaging with both LiDAR on drones,\
\
Angie:    as well as photogrammetry.\
\
Angie:    Christoph isn't here, but he and I are working on developing\
\
Angie:    an intermediate range recognized drone for deployment at\
\
Angie:    stratovolcanoes with a company called Black Swift technologies.\
\
Angie:    This is their S2 platform.\
\
Angie:    It has interchangeable nose cones.\
\
Angie:    Right now, they're developing nose cone that includes gas sensors for us,\
\
Angie:    as well as a nadir-looking camera for photogrammetry and\
\
Angie:    then a real time video feed off off the front.\
\
Angie:    The great thing about this is we can use it at stratovolcanoes.\
\
Angie:    It can fly over 20 kilometers from deployment,\
\
Angie:    it can stay in the air for an hour and a half,\
\
Angie:    and I think that this will really revolutionize a way\
\
Angie:    that we can monitor kind of more dangerous volcanoes.\
\
Angie:    Kilauea  was a relative exception because we can fly from\
\
Angie:    on top of the volcano whereas other ones we know we need to keep a safe distance.\
\
Angie:    We were hoping to deploy this new drone at Makushin in Alaska this summer,\
\
Angie:    but we're not sure where that stands given COVID-19 restrictions,\
\
Angie:    but hopefully we can do that soon.\
\
Angie:    But we also want to continue work with our VDAP partners,\
\
Angie:    applying drone technology, to monitoring their active volcanoes,\
\
Angie:    as well as conducting various workshops.\
\
Angie:    Bottom right is a workshop that I led in Chile in March of this year,\
\
Angie:    where we were using UAS to map\
\
Angie:    some volcanic deposits and I have used them\
\
Angie:    extensively with our partners in Indonesia as well as Peru.\