Tag: UD

Atlantic sturgeon arriving earlier in the mid-Atlantic

The unusually warm conditions in the winter and spring of 2012 have resulted in water temperatures up to 3°C warmer than the previous 3 years resulting in comparable Atlantic sturgeon catches off the coast of Delaware occurring 3 weeks earlier than past sampling efforts.  During sampling events for Atlantic sturgeon we have also documented sand tiger sharks arriving off the coast of Delaware in late-March, a full month earlier than documented in previous seasons.

My research, conducted jointly with Dewayne Fox at Delaware State University and Matt Oliver at the University of Delaware, is focused on coastal movements and habitat use of adult Atlantic sturgeon during the marine phase of their life history.  By utilizing acoustic biotelemetry on both traditional fixed array platforms as well as developing mobile array platforms coupled with Mid-Atlantic Regional Association for Coastal Ocean Observing Systems (MARACOOS) I am going to model Atlantic sturgeon distributions in a dynamic coastal marine environment.  This research is particularly relevant given the recent protection of Atlantic sturgeon under the Endangered Species Act.  Determining factors influencing Atlantic sturgeon movements and distributions during their marine migrations will enable dynamic management strategies to reduce mortalities as well as impacts to commercial fisheries, dredging efforts, and vessel traffic.  In addition to allowing for dynamic management strategies the development of models for adult Atlantic sturgeon movements and distributions in relation to dynamic environmental conditions will illustrate how changing environmental conditions are going to impact this Endangered Species moving forward.

Graduate student Matt Breece with a recently telemetered female Atlantic sturgeon off the coast of Delaware

Utilizing New Tagging Technology to Characterize Sand Tiger Shark Habitat

Sand tiger sharks are large bottom dwelling sharks found in the coastal waters of the Eastern North Atlantic, and are known to frequent the Delaware Bay in the summer months. Sand tiger shark populations are currently in danger of over exploitation because they are slow growing and have extremely low birth rates. While we know that the sharks are found within the Delaware Bay during summer months, little is known about their movements during the rest of the year, or what oceanographic conditions limit their spatial extent. There is evidence that these sharks make large coastal migratory movements along the Eastern Seaboard. This makes habitat characterization difficult because the sharks travel throughout such a large area. It is important for managers to know the areas of intensive use by the sharks, and the species assemblages within those areas, in order to protect these apex predators.

Dr. Matthew Oliver and Danielle Haulsee with a sand tiger shark caught in the Delaware Bay.

Our project, a collaboration among Delaware State University’s Dewayne Fox and the University of Delaware’s Matthew Oliver and Danielle Haulsee, will document and characterize the movements of sand tiger sharks, their habitat preferences, and the community assemblages they encounter using new and innovative electronic tagging technology. Sand tiger shark movements will be recorded using passive telemetry in addition to pop-off satellite archival tags. We will also deploy a new type of tagging technology, which acts as a mobile receiver, and will record any encounters with other sharks, fish or other marine animals that have been tagged with acoustic tags. We will then use satellite and remotely sensed data resources from the Mid-Atlantic Regional Association for Coastal Ocean Observing Systems (MARACOOS) to characterize and model the habitats and oceanographic conditions used by sand tiger sharks. This study will give managers a better understanding of the spatiotemporal patterns in sand tiger shark movements along the East Coast, as well as inform management decisions regarding sand tiger shark habitat utilization.

Predicting Sea Surface Salinity from Space

The simplest definition of salinity is how salty the ocean is. Easy enough, right? Why is this basic property of the ocean so important to oceanographers? Well, along with the temperature of the water, the salinity determines how dense it is. The density of the water factors into how it circulates and mixes…or doesn’t mix. Mixing distributes nutrients allowing phytoplankton (and the rest of the food web) to thrive. Globally, salinity affects ocean circulation and can help us understand the planet’s water cycle. Global ocean circulation distributes heat around the planet which affects the climate. Climate change is important to oceanographers; therefore, salinity is important to oceanographers.

Spring Salinity Climatology for the Chesapeake

Spring Salinity Climatology for the Chesapeake

Salinity doesn’t vary that much in the open ocean, but it has a wide range in the coastal ocean. The coast is where fresh water from rivers and salt water in the ocean mix. Measurements of salinity along the coast help us understand the complex mixing between fresh and salty water and how this affects the local biology, physics, and chemistry of the seawater. However, the scope of our measurements is very small. Salinity data is collected by instruments on ships, moorings, and more recently underwater vehicles such as gliders. While these measurements are trusted to be very accurate, their spatial and temporal resolution leaves much to be desired when compared to say daily sea surface temperature estimated from a satellite in space.

So, why can’t we just measure salinity from a satellite?Well, it’s not as simple, but it is possible. NASA’s Aquarius mission http://aquarius.nasa.gov/ which was launched this past August is taking advantage of a set of three advanced radiometers that are sensitive to salinity (1.413 GHz; L-band) and a scatterometer that corrects for the ocean’s surface roughness. With this they plan on measuring global salinity with a relative accuracy of 0.2 psu and a resolution of 150 km. This will provide a tremendous amount of insight on global ocean circulation, the water cycle, and climate change. This is great new for understanding global salinity changes. What about coastal salinity? What if I wanted to know the salinity in the Chesapeake Bay? That’s much smaller than 150 km.

That’s where my project comes in. It involves NASA’s MODIS-Aqua satellite (conveniently already in orbit: http://modis.gsfc.nasa.gov/), ocean color, and a basic understanding of the hydrography of the coastal Mid-Atlantic Ocean. Here’s how it works: we already know a few things about the color of the ocean, that is, the sunlight reflecting back from the ocean measured by the MODIS-Aqua satellite. We know enough that we can estimate the concentration of the photosynthetic pigment chlorophyll-a. So not only can we see temperature from space, but we can estimate chlorophyll-a concentrations too! Anyway, there are other things in the water that absorb light besides phytoplankton and alter the colors we measure from a satellite.

Spring Salinity Climatology for the Mid-Atlantic

Spring Salinity Climatology for the Mid-Atlantic

We group these other things into a category called colored dissolved organic material or CDOM. CDOM is non-living detritus in the water that either washes off from land or is generated biologically. It absorbs light in the ultraviolet and blue wavelengths, so it’s detectable from satellites. In coastal areas especially, its main source of production is runoff from land. So, CDOM originates from land and we can see a signal of it from satellites that measure color. What’s that have to do with salinity?

You may have already guessed it, but water from land is fresh. So, water in the coastal ocean that is high in CDOM should be fresher than surrounding low CDOM water. Now we have a basic understanding of the hydrography of the coastal Mid-Atlantic Ocean, how it relates to ocean color, and why we need the MODIS-Aqua satellite to measure it. So, I compiled a lot of salinity data from ships (over 2 million data points) in the Mid-Atlantic coastal region (Chesapeake, Delaware, and Hudson estuaries) and matched it with satellite data from the MODIS-Aqua satellite in space and time. Now I have a dataset that contains ocean color and salinity. Using a non-linear fitting technique, I produced an algorithm that can predict what the salinity of the water should be given a certain spectral reflectance. I made a few of these algorithms in the Mid-Atlantic, one specifically for the Chesapeake Bay. It has an error of ±1.72 psu and a resolution of 1 km. This isn’t too bad considering the range in salinity in the Chesapeake is from 0-35 psu, but of course there’s always room for improvement. Even so, this is an important first step for coastal remote sensing of salinity. An algorithm like this can be used to estimate salinity data on the same time and space scale as sea surface temperature. That’s pretty useful. The folks over at the NOAA coastwatch east coast node thought so too. They took my model for the Chesapeake Bay and are now producing experimental near-real time salinity images for the area. The images can be found here: http://coastwatch.chesapeakebay.noaa.gov/cb_salinity.html. They will test the algorithm to see if it is something they want to use

Climatologies of salinity for all of my models can be downloaded here: http://modata.ceoe.udel.edu/dev/egeiger/salinity_climatologies/.

I view this project as an overall support of the NASA Aquarius mission by providing high resolution coastal salinity estimates that are rooted in in situ observations. I hope this information proves to be useful for coastal ocean modeling and understanding the complex process that effect the important resource that is our coasts.

Demobilization and Remobilization of the Hugh R Sharp

Summer is an especially busy time for research vessels. The UNOLS fleet is making increasing use of containerized portable lab vans to shave some time and effort off of offloading the science party from one cruise and loading up the next mission and their gear. They also increase the flexibility of the research vessels by giving them the option to add additional science capabilities and facilities to vessel users. Options include adding:

  • Dry Labs
  • Wet Labs
  • Isotope Labs
  • Clean Labs
  • Cold Labs
  • Additional Berthing

This is a time lapse that we shot of the RV Hugh R Sharp returning from a multi-week scallop survey, unloading one lab van and then loading two more fresh ones before fueling up (both diesel and food) and departing on the next mission. Enjoy!

It’s all about the E-Lec-Tricity

We had a gentleman named Matthew Vest come to the GVis lab the other day to show off his do-it-yourself creation. He was looking for information on whether he might be able to showcase it at the upcoming Coast Day 2011 event that happens each year on the second Sunday of October here at the Hugh R Sharp campus in Lewes.

Matthew has done something that most of us dream about doing, something many of us say we’re going to do, and that same something that most of us never get off our duffs and actually do. He has taken a 1985 Chevy S-10 truck, removed the gasoline engine and tank, and replaced them with an electric drive motor (from a fork lift he says) and a bed full of 6 volt lead acid batteries (aka ‘golf cart batteries’ – 24 in total). The conversion took him about 2 years to complete and cost approximately $10,000 dollars but now he is the proud owner (and creator) of an all-electric vehicle that will to approximately 40-60 miles on a charge.

Matthew Vest' Electric Truck

Matthew Vest’ Electric Truck

Matthew went out of his way to select “Deka” batteries to power his creation, which he says are 100% recyclable. Each of the 24 batteries weighs in at 60 lbs, for a total battery weight of about 1450lbs. These batteries are wired in series to generate the 144 volts DC that power the Warp-9 electric motor that replaces the gas engine. There is one 12 volt battery which is used to power the stock lights, wipers and horn.

Chevy S10 - Batteries in the Back

Chevy S10 – Batteries in the Back

A Curtis 1231c controller is like the brains of the truck, controlling the power flow. A Zivan ‘smart charger’, which runs on standard 110v, sits behind the driver seat. When fully discharged, the batteries take about 10-12 hours to recharge. The only sound that the truck makes when it is running is the sound of the add-on vacuum pump that is also under the hood. It creates the vacuum that assists with the stock braking system of the truck.

Chevy S10 - Under the Hood

Chevy S10 – Under the Hood

Matthew is hoping to touch base with some of the researchers at UD that are involved in the V2G or “Vehicle To Grid” project so that he can assess whether his S10 can also be integrated with the power grid. For more information on V2G and GIEV’s (Grid Integrated Electric Vehicles) you can read more on the Q&A section on the V2G site.

We asked Matt how he got started with the project and he said it just took some research online, a couple of “how to retrofit your gas vehicle into an electric vehicle” books, and some very helpful people on a few of the EV forums. We salute Matt for what turned out to be an excellent EV refit and for his consideration of the environment when he selected the batteries and materials for his electric vehicle project. Well done!

 

Hurricane Katia Footprints

The ORB Lab was having a meeting in the GVis Lab this week and, as usual, the East Coast US 8-Day Averaged Sea Surface Temperature overlay was up on the screens. Dr. Oliver pointed to the screen and noted that there was a path cutting across the Gulf Stream that was cooler than usual and that it was probably due to upwelling and mixing from hurricane Katia. Sure enough, we loaded up a layer showing Katia’s track and they lined up.

Katia SST Trail

Katia SST Trail

We then checked to see if there was anything noticeable on the East Coast US 8-Day Average Chlorophyll layer and you can see what appears to be a slight bloom in chlorophyll along the track as well (slightly lighter blue).

Katia Cholorophyll Trail

Katia Cholorophyll Trail

Another neat view is the markedly cooler water that you flowing into the bays from the increased river discharge that resulted from the large amounts of rain dropped by hurricane Katia and tropical storm Lee as they passed through.

Cold river water 20110913

Cold river water 20110913

These layers and several others are processed and uploaded daily and made available via the Orb Lab website in the Public Access section. They are exposed via Google Maps interfaces as well as Google Earth embedded views and linkable KMZ file formats. Neat stuff!

A Wind Turbine Experience

Luckily Blaise Sheridan is not afraid of heights, as he climbs up the UD 2-megawatt wind turbine for the second time. With his Master’s thesis revolving around wind energy, he is one of only four people from UD certified to climb the turbine. Although there is an elevator (more technically termed a personnel or ‘man’ lift), it can only be used by those who take a more intensive 4-day training course. Instead, a 2-day Fall Protection/Competent Climber class was taken by two facilities employees (Don Smith and Rodney McGee), as well as two UD students (Blaise and DeAnna Sewell). With this course under their belt, they can climb the ladder to the top of the 256-foot-tall-turbine. For their safety, they are always connected to a guide wire that clips onto the cable grab of each climbers harness. The cable ensures that if a climber falls they will only drop less than a foot.

 This goal of this trip was to string up 3 cables to install bat microphones. The microphones will allow researchers to see how often bats pass around the turbine. This anticipated one-day job ended up taking about 2.5 days due to lightning and the large amount of on site planning that needed to take place. With the help of a Gamesa contractor, Blaise and Rodney were able to install the research equipment while the contractor performed routine maintenance and provided his expert guidance.

The turbine is currently producing more electricity than projected, although how much more is still being studied. On average, it produces more energy than the university needs, which makes the excess available to the town!

Inside the nacelle, the bus sized structure on the top of the tower where all the interesting mechanical and electrical components are housed, Blaise notes,  “It must be at least 120 degrees” from the waste heat given off by the electrical transformers, not to mention all the gearboxes, friction and the fact that heat rises up the turbine. But, outside, on top of the nacelle, there’s enough airflow to cool you off! Blaise admits it can be very tiring to climb but the incredible view from the top is worth it. He discloses his favorite part is to watch the wake off the boats coming into Roosevelt Inlet. With the hope of additional renewable energy options in the future,  “It’s still very novel for a university to have this turbine and its been a once in a life time experience…one to check off the bucket list.  Not to mention it’s a great bar story.”

Timelapse of a Day in the ORB Lab’s GVis Room

I was showing the students how to operate the “birdcam” so they can use it to record a series of stills to create a time lapse video of an upcoming research cruise on the RV Hugh R Sharp. We left the birdcam in the corner and let it click away all day, shooting a new still every minute and the video above is the resulting masterpiece. It is embedded from “The UD ORB Lab” channel on YouTube.

You can learn more about the “birdcam” in a previous post about “Timelapse Video on the Cheap“. The GVis Room pictured above is the “Global Visualization Room” that was described in the post “How to Construct a Global Visualization Lab“.

Thanks to the ORB Lab crew for sharing!

 

Conch Reef Survey for NASA’s NEEMO 15 Project

Dr. Art Trembanis’ Coastal Sediments, Hydrodynamics & Engineering Lab (CSHEL) has been pretty busy lately. Not long ago I did a post about the prototype sub-bottom profiler section that he added to his Autonomous Underwater Vehicle (AUV) (see: Sub-Bottom Profiling using an AUV). I was down at the NASATweetup for the Endeavour (STS-134) launch not long ago and I got chatting with some folks from NASA’s Open Goverment Initative about the NEEMO 15 project (NEEMO stands for “NASA Extreme Environment Mission Operations“) and we discussed UD’s involvement.

It takes a village of roboticists to run a successful AUV campaign

It takes a village of roboticists to run a successful AUV campaign

When I emailed Dr. Trembanis upon my return to Delaware, he emailed me back with instructions to browse to UNCW’s Life Support Buoy live webcam above the Aquarius Reef Base. Sure enough, he was there aboard the RV George F. Bond monitoring his Gavia Scientific AUV as it acoustically mapped the Conch Reef around the Aquarius as a precursor robotic mission for NEEMO 15.

Go Pro Hero Attached to the AUV

Go Pro Hero Attached to the AUV

Here is video footage shot by an off-the-shelf HD Go Pro Hero digital video camera that was attached to the AUV:

httpv://www.youtube.com/watch?v=8n3nR9TaVGo

The mapping mission ran for 4 days and covered approximately 100km, resulting in about 15Gigabytes of raw data. Here’s an overview map of the mission.

Aquarius NEEMO 15 precursor survey

Aquarius NEEMO 15 precursor survey

Many thanks to Dr. Trembanis for the video and imagery to go along with the story. Be sure to visit NASA’s NEEMO site to learn more about the mission and what’s to come. Visit the CSHEL site to learn more about the research that’s going on there and to see other cool video and image products that they’re producing.

Sub-Bottom Profiling using an AUV

I was minding my own business, walking between Smith Lab and Cannon Lab buildings when what to my wandering eyes should appear but a reeeallly long stretched out Gavia Scientific AUV. My geek radar started going off and I just HAD to investigate exactly what was inside these newly milled sections of hull.

Gavia Scientific AUV with a recent addition

Gavia Scientific AUV with a recent addition

I invited myself into the lab and started asking some questions. It turns out that these new sections contain a prototype Teledyne Benthos Chirp III sub-bottom profiler that was specially designed to integrate with an AUV. Dr. Art Trembanis’ CShel lab and Val Schmidt from the University of New Hampshire’s Center for Coastal and Ocean Mapping were working with UTEC Survey Inc. to successfully integrate and test this new addition to the AUV’s sensor lineup. I cornered Nick Jarvies from UTEC and he gave me the run-down on the new addition (thanks Nick!):

httpv://www.youtube.com/watch?v=fQkWAhaFcsk

Sample SBPWhat is a “sub-bottom profiler” you ask? Per the Wikipedia entry, it is a “powerful low frequency echo-sounder…developed for providing profiles of the upper layers” of the ocean floor. In the case of the Chirp III, probably in the range of 10-20kHz. Per Dr. Trembanis “Data is stored in an onboard Compact Flash card in an industry standard SEG-Y format.  The advantage of a chirp signal over a single frequency output is that through chirp demodulation of the returning signal one can get a better compromise between penetration and resolution.  The lower the frequency the greater the penetration but the less the resolution (and vice versa for high frequency) so a chirp signal which modulates from a low to high frequency provides penetration and resolution.  All of this depends to a great degree on the kind of bottom material one is trying to penetrate.”

Internal view of the Benthos Chirp III AUV SBP

Internal view of the Benthos Chirp III AUV SBP

The advantages of an AUV-based sub-bottom profiler (also per Art Trembanis) are:

  • We remove lots of water column data that would normally be unwanted and has to be removed/ignored from the record.
  • Because we can precisely follow the terrain near the bed or hold a constant depth well below the surface we can remove/diminish effects of waves that cause a ship to bob up and down.
  • We are able to do higher resolution characterization of the subsurface in greater water depths since otherwise from a surface ship you would have to use a lower frequency system to penetrate through the water column.
  • Because of the precise navigation of the AUV we can get very tight line spacing and precision following of features (i.e. pipeline routes) which allows us to provide better data more efficiently.

Thanks to everybody for taking time to talk on camera and for answering my questions!

Outdoor Webcam 101

Some time ago I was asked what would it take to get a live webcam feed of the osprey nest next to our Marine Operations Building. We have an osprey couple – Ricky & Lucy – and people love to check in on them throughout the summer months when they come home to Lewes.

Ricky & Lucy
Ricky & Lucy

I thought I’d share the software and hardware lineup that I selected to do the job and explain some of my choices. The equipment I ended up ordering was:

  • Sony SNC-RZ30N PTZ (pan + tilt + zoom) IP Webcam (~$1,100)
  • Dotworkz D2 Outdoor Enclosure with heater/blower (~$500)
  • Videolarm APM3 Pole Mount Bracket (PDF) (~$60)
  • WebcamXP network camera monitoring, streaming & recording software (~$99)
  • Some sort of intermediary computer to run the WebcamXP software
  • 100′ Outdoor Extension/Power Cord
  • 100′ Underground Double-shielded Cat5 Network Cable
  • Uninterruptable Power Supply (UPS)
  • Desiccant packs, Velcro Tape & a Plastic Container
  • RainX and Marine Silicone RTV

Why I picked what I did

We wanted to mount the webcam on an existing antenna tower next to the Marine Operations Building, thus the pole mount bracket. The webcam needed to have a short shopping list of features – including:

  • Stand-alone operation and a physical network jack for IP (network) streaming – we didn’t want to potentially impact building wifi performance and we just wanted to be able to run a network line up to it (no USB webcams need apply). Doing this with a network cable tied into our building switch meant that the packets the camera generated would not negatively impact the rest of the users.
  • It needed to have sufficient intelligence that we could remotely log into it to position it and/or program preset camera stops and zoom factors.
  • A healthy optical zoom so we could zoom up close on the nest for an up-close experience.
  • Image stabilization built-in so that when we did zoom in, any sway or vibration in the antenna pole wouldn’t give us a jittery image.
  • FTP/FTPS functionality – where you can have the webcam automagically FTP a still frame to an outside server at a user-defined interval. We used this feature to amass the still shots that we used in our (award winning) time lapse videos for the Lewes wind turbine construction. (We stopped the camera from moving for the 2-3 weeks it took to complete construction.)

We selected the Sony SNC-RZ30N for the job, but before you go out hunting for one, they seem to have been discontinued. In its stead now is the SNC-RZ25N (slightly lesser 18x optical zoom than the 30N) and its replacement, the SNC-RZ50N (26x optical zoom) which does both motion JPEG and H.264 streaming. The SNC-RZ30N camera has its own built-in web server so we could control it via a web browser. It has a 25x optical zoom so we can get up close and personal with the osprey nest. Don’t be fooled by some webcams which tout a zoom without specifying that it’s an optical zoom. If it doesn’t say “optical zoom” it’s most likely a digital zoom, meaning a lower resolution subset of the total number of pixels the camera can capture.

The SNC-RZ30N supports up to 16 presets, which allows you to position the camera where you want it pointed at the zoom factor you want, and to save a “preset”. You can then have the camera cycle itself through the various presets at a user-specified panning speed, stopping at each stop for a user-specified amount of time. Quite handy when we are cycling the webcam to look at various points of interest on campus, and even handier for removing the osprey nest preset from the mix when the ospreys head south for the winter.

Sony web interface

Sony web interface

As you can see, the webcam is pretty close to the ocean, so we needed to find an enclosure for it that could:

  • Survive a salty marine environment
  • Remain water-tight
  • Provide space for desiccant packs to remove any excess moisture so that the inside of the enclosure didn’t fog up on cold mornings
  • Provide an automatic heating of the enclosure on cold mornings to prevent frosting up of the outside dome
  • Provide power for the camera inside

We chose the Dotworkz D2 Enclosure with optional heater. It has two sealable penetrations that allowed us to get power to the unit by cutting off the end of a heavy-duty outdoor extension cord, tinning the tips and tightening up the screws onto it to power the power supply and heater inside (green circle). The other end of the extension cord is plugged into a UPS/Surge Protector in the radio room below since the webcam is strapped to a huge metal pole sticking up into the (sometimes lightning filled) sky. An underground double shielded network cable was run up the tower and inserted through the second penetration – after which we crimped a RJ45 end onto it and simply plugged it into the back of the webcam. The power supply came with an end that was already compatible with the power connector on the back of the webcam (orange circle) so powering the camera was a cinch.

Enclosure interior

Enclosure interior

The enclosure also came with a magic universal mounting bracket and stand-offs of various heights to ensure you can position a compatible webcam at the right height to see out the bottom dome.

Mounting plate stand offs

Mounting plate stand offs

Here is a picture of the connectors on the back of the webcam:

Sony back

Sony back

We attached the camera as high as we could and still reach it with a scissor lift that facilities owns. The first time we installed the enclosure, we had it mounted by the pole climbers that were doing maintenance on the antennas at the top of the tower (no way I could ever do that – waaayyy too high up). We relied on the seal built into the enclosure to handle sealing out the moisture, but unfortunately it slowly started amassing some moisture inside which started pooling up inside the dome over several months. Since we didn’t have the requisite climbing gear to climb the tower, we ended up having the tower climbers move the enclosure down the pole just high enough for us to reach with the scissor lift when they came back again. We didn’t take any chances this time. We went along the exterior seam and penetrations with some marine RTV (silicone sealant) and I used some velcro tape to secure a plastic container filled with desiccant packs on top of the black mounting plate to keep the inside as dry as possible. The velcro would keep any tower vibrations from storms and the like from working the desiccant packs over the edge and down onto the dome.

Webcam on tower

Webcam on tower

We made sure to loop some additional network and power cable up the tower just in case we needed to move the enclosure to a different height or to another side of the tower. Make sure to make a “drip loop” with any cables that dips down from the enclosure and then back up away from it. This keeps water from flowing down the cable and running against the penetration, thus minimizing the likelihood of water making its way into the housing. Remember that the cables are exposed to the elements, which includes ultraviolet radiation (UV from sunlight) which can break down most plastics and vinyl cable sheathing. We selected an extension cord which listed UV resistance and selected an outdoor network cable to stave off the UV damage to the cables.

Webcam on tower closeup

Webcam on tower closeup

One last treatment I did was to apply Rain-X to the outside of the dome. It’s like a wax coating for glass that makes water bead up and roll down the dome rather than stick to the outside and fog up your view.

WebcamXP

One last topic I’d like to cover is the use of WebcamXP as a bridge between the webcam and the outside world. The problem with most webcams is that of security and scale. The internal web server in most webcams can handle about 25-50 simultaneous users. If you have more than that number, attempts to view the feed by users 51 and up will fail. To overcome this limitation, we purchased WebcamXP as an intermediary. The software installs on a desktop or server and it makes a connection to the webcam and handles the task of streaming it to the web server that you’re embedding the feed on. By acting as the intermediary, WebcamXP offloads the streaming load from the webcam. In our case we embed a Flash SWF file on the external webserver that gets its stream from WebcamXP.

The second issue that you run into with many webcams is that of security. They have some basic security built-in, but in order to stream the video from many of them, you have to expose the ability to control and position the webcam to the end-user. The last thing we wanted was for random users repositioning the webcam. Our solution was to give the webcam an internal IP address that was not accessible from outside our border routers. The system running WebcamXP was given a publicly accessible IP address and an internal IP address so that it could access the webcam video stream and serve it up externally.

Other nice features of the software are:

  • Watermarks – the software allows you to embed a watermark image over your video stream, thus branding your video with your logo and/or text.
  • Ability to expose the video stream via Java, Javascript or a Flash client.
  • Ability to handle multiple IP webcams simultaneously. If you want to grow the number of webcams you want to expose, you would only need one system running WebcamXP to stream the feeds from multiple webcams simultaneously.
  • A free version, which can handle a single webcam. This allows you to kick the tires and make sure the software does what you want before you buy. (Note, the free version does not allow watermarking your logo on the video stream)

I had initially looked into using either Silverlight and/or the IIS Streaming Server to handle this roll, but it was early in their development when we set the webcam up and it was more expedient to use WebcamXP. I’d still like to look into having our actual web server do the work of connecting to the internal webcam and handle streaming the content using Silverlight or some other non-Flash mechanism. If you have some feedback as to how to accomplish this, I’m all ears. I think it would make a much more flexible mechanism to handle the various browsers (including mobile, iPad etc.) that are coming online.

Thanks for enduring the long post and please feel free to comment if you can think of things I missed or have any suggestions on how to improve things.

Trip to Penguin Colony on Biscoe Point

Folks seem to like penguins….so much so that we even made the front page of the University of Delaware website! Hurray! This shot is of the Adélie penguin colony on Humble Island. We had just deployed a satellite transmitter on one of the birds so we would know where to send the underwater robots (Gliders and REMUS’s).

University of Delaware Main Page

Remnants of the storm remain in the area and wind gusts are keeping the science boats at station today. Nevertheless, we did have a break in the clouds and the sun came out. The warm sun made the Gamage glacier very active and I happened to get a great video of calving. Right place, right time.

httpv://www.youtube.com/watch?v=vANNYwc2d_k

We headed out to Biscoe Point to deploy another satellite transmitter on a penguin. The plan was remove the transmitter from a Gentoo penguin which had been at Biscoe Point since mid-night. The challenge is to find the tagged bird amongst the rest! On the way, a large amount of brash ice had surrounded Biscoe Point, so we had 1-2km if slow travel through the ice. Marc Travers (our boat driver and expert birder) did an excellent job snaking in between the large chunks. Outboard motors and large chunks of brash ice don’t mix well. Hitting a large piece of ice can leave you on a boat with a busted motor. That is why we carry an extra motor in every boat.

httpv://www.youtube.com/watch?v=URLwiXtnZk0

When we arrived at Biscoe Pt. we found that an Elephant Seal had climbed into one of the Gentoo Penguin nesting areas. If the penguin chicks are too young or unguarded by its parents, they can be easily crushed by these massive seals.

Southern Elephant Seal in Gentoo Penguin Colony overshadowed by Mt. William

Luckily it looked like the Gentoo chicks were old enough to avoid it. Occasionally a Gentoo adult would peck at the Elephant Seal’s thick blubber, but the giant beast didn’t seem to be bothered by it at all.  We made our way around the Gentoo colony looking for our tagged bird. She happened to be perched right on a rock preening herself where we could see her plain as day. The birders quickly removed the tag and she went back to her nest.

Penguin Chick Eaten by Skua Birds

Elephant Seals aside, the biggest threat to the chicks are Skua’s. These are aggressive scavenger birds swoop down and grab chicks right from their nests and make a meal out of them. There was plenty of evidence at Biscoe Pt. that the Skua birds had been active here.

Still, even with the ever present Skua, there were plenty of Gentoo chicks that were starting to look more and more like their parents. They are are starting to get their adult feathers. Their feathers are not waterproof yet, but they will be soon.

Gentoo Chick with Parent at Biscoe Pt.

The next step was downloading the dive information from the tag. This data will help us understand how deep the penguins are feeding. The dive data will help us properly analyze the data coming from the underwater Gliders and REMUS vehicles. The Birders are able to download and ready the tag for its next deployment in just a few minutes with a laptop computer in the field. These are amazing little tags.

httpv://www.youtube.com/watch?v=v-lLPVhjnMY

We walked around a small bay to the neighboring Adélie Penguin colony and were able to quickly identify an Adélie penguin that would be good for carrying our satellite

Adélie Penguin packed with a satellite transmitter.

transmitter. She was quickly tagged and released back to her nest. Her two grey puffy  chicks are just to her right. We will be watching the satellite data closely to find out where she is eating. Then, we will send our underwater robots to sample that section of ocean.  In a few days the Birders will head to Biscoe Pt. again to retrieve the tag, and thank her for her contribution to science.

Antarctic Storm Moves In

Our streak of excellent weather has officially come to an end with a large low pressure system in the Drake Passage.

Storm moves into Palmer Station

The weather was even tough tough for the ever-working “birders” who were going to deploy a few satellite tags on penguins today. REMUS missions are cancelled for the day. That might be good since one sprung a leak on a mission yesterday. Only the gliders are out….which makes gliders an awesome platform for ocean science when the weather gets a bit “snotty”. They don’t complain and don’t get sea-sick. The “Blue-Hen” continues is mission mapping the foraging locations of penguins when even the penguins are too scared to go out! That means I get to stay home and peel garlic (very necessary for all the amazing food here).

Garlic….it’s like the best thing you can eat when it is windy

Saturday is also the day we all clean the station and have a station meeting. I got to help clean the kitchen today. That was really nice because I totally miss cleaning the kitchen at home (no, not really). We also learned that hiking on the Gamage glacier behind the station is more restricted after a new crevasse opened up. Funny story about that…..Mark Moline found it by falling into the crack. He was fine, but it was a bit un-nerving. The GSAR (Glacial Search and Rescue) team changed the boundaries after they went and uncovered the full extent of “Mark’s Crack”.

The bad weather lets us do a bit of data analysis on where the penguins are foraging. The penguins seem to be keying off of the deep canyon off of Palmer station. This has been a working hypothesis from the “birders”

Finally, I’ll leave you with an awesome moon-rise over the Gamage Glacier. Pretty awesome sight.

Moonrise over Gamage Glacier

Penguins, AUV’s, Satellites: together at last

Adélie Penguin Rookery

Adélie Penguin Rookery on Humble Island

Satellite tagged Adelie Penguin

Satellite tagged Adelie Penguin

Penguin swimming tracks near Palmer Station

Penguin swimming tracks near Palmer Station

Ballasting the Glider (Blue Hen)

Ballasting the Glider (Blue Hen)

Is it possible to follow penguins from space to understand where and how they are feeding in Antarctica? Absolutely!..but not without an excellent team from University of Delaware, Rutgers University, Polar Oceans Research Group, and Cal Poly San Luis Obispo. The sequence starts with the “Birders”. The “Birders” are from Polar Ocean Research and they have been studying penguins in the West Antarctic Peninsula for years. The “Birders”, headed by Bill and Donna Fraser, head out to local rookeries to identify good penguins to tag with satellite transmitters. Finding the right breeding pair is key. The pair should have two chicks with both parents still around. Some chicks only have one parent, probably because one parent was killed by a Leopard Seal. We want to choose one of the parents, because we are pretty certain they will return to their chicks to feed them. This also helps in recovering the transmitter. If the bird does not return, the transmitter comes off during their natural annual molt cycle. Once a penguin is selected, it is gently fitted with a satellite transmitter. Special waterproof tape is used to connect the transmitter to the thick feathers on the back of the penguin. The penguins are remarkably calm during the process.  Once the tag is attached, the penguin is released back to its nest. The next part of the sequence is for the birds. The penguins head out to feed on krill and small fish in the area. Their tags relay their position information to ARGOS satellites and we get nightly updates. The Birders pass on their data to me nightly, and I filter and map the penguin tracks. I put them into Google Earth, so we can see where the penguins have been feeding. Then, through the magic of mathematics, we turn their tracks into predicted penguin densities. Based on these densities, we plan our AUV missions to intersect with the feeding penguins (Slocum Electric Gliders and REMUS AUV’s).  The first priority is to make sure the AUV’s are ballasted correctly. This means that they need to be trimmed with weights just right so they travel correctly under the water. We use small balances and scales to get the weight of the vehicle just right, then put them into ballasting tanks to make sure we did it correctly. The vehicles should hold steady just under the surface of the water.

Getting ready for the launch of the "Blue Hen"

Getting ready for the launch of the "Blue Hen" (M. Oliver and K. Coleman)

Once we have a planned mission, we head out in small zodiacs from the station to a pre-determined point. For the Gliders, we call mission control at Rutgers University (Dave, Chip, John) and let them know a glider will be in the water shortly. Once it is in, control of the glider is accomplished via satellite telephone directly to the glider. The glider calls in and reports data and position to mission control. We can see the data coming in live over the web, and in Google Earth as we navigate the vehicle to where the penguins are feeding. The gliders move by changing their ballast, which allows them to glide up and down in the water while their wings give them forward momentum. They “fly” about 0.5mph for weeks at a time!

Mark Moline with REMUS's

Mark Moline with REMUS's

In contrast to the Gliders, the REMUS vehicles are very fast and are designed for shorter, 1 day missions. Daily missions are planned around the penguin foraging locations. The Cal Poly Group (Mark Moline and Ian Robbins) have been launching 2 Remus Vehicles per day to map areas the gliders can’t get too. Like the gliders, these vehicles call back via iridium to let us know how they are doing in their mission.

MODIS Chlorophyll, Penguins, and Gliders

Glider Dances around Adélie Penguin Tracks in a sea of chlorophyll

Finally, we are getting satellite support from my lab at U.D. Erick, Megan and Danielle have been processing temperature and chlorophyll maps in near-real time to support our sampling efforts, as well as AUV operations up and down the West Antarctic Peninsula. Just today, we saw that the penguins in Avian Island (south by a few hundred miles) have been keying off of a chlorophyll front. RU05 was deployed by the L. M. Gould and will be recovered soon. All in all, it is a pretty awesome mission to track these penguins from space and AUV’s. We will see how the season develops!

Note: I will be uploading photos and videos to the ORB Lab Facebook page throughout my stay in Antarctica. Be sure to check there for my latest updates.

How to Construct a Global Visualization Lab

My apologies for how long it took to get this up. I promised our colleagues at the Xiamen University that I’d put up the complete specs for the Global Visualization room – a component of Dr. Matt Oliver’s ORB Lab and the pesky day job kept getting in the way.

Panorama Fish Eye LensI originally tried a video walk-through of the GVis Lab but it ended up being a lot of panning and zooming around, which I didn’t really care for. Instead I got to try out a fancy digital camera with a 180 degree fish-eye lens the other morning, which I used to shoot three shots of the room 120 degrees apart from each other. I used a software package called 3DVista Show to stitch the fish-eye pictures into a panorama image, which I uploaded to a free online hosted tour on their site. Once I got through uploading the image, the service provided an iFrame string that I included in the post to embed the panorama project. Be sure to click the full-screen icon (top right-hand arrow next to the question mark) to see the panorama a little better.

As you pan around the room, you’ll see the major components of the lab, which are:

The Dell Precision T7500 workstation was selected because it was one of the few systems that was capable of handling (2) PCIe x16 graphics cards simultaneously. We started with one graphics card with the expectation that it could handle the video workload, but wanted the option to add another graphics card in SLI mode to boost graphics performance. So far we haven’t needed a second video card as everything runs quite smooth under Windows 7 x64 as the base operating system and running Google Earth Professional.

The nVidia graphics card has two DVI outputs. One output is fed into the VWBox 133A video splitter, which spreads the 4300×2100 signal across the (9) monitors in the 3×3 monitor array. The VWBox also allows us to “subtract out” the bezel, which eliminates a few lines of video where the bezels are – making for no stepping in diagonal lines or graphics. The 460UX-2 Samsung monitors are all 1920×1080 (1080p) monitors with an 11mm bezel on all four sides. This is the smallest bezel monitor that was available when we built the wall. For Google Earth and other high-resolution work, the display is fantastic, however a second monitor was added to display lower-resolution content at a larger size and for Powerpoint presentations and the like. As small as the bezels are, they may cause some readability problems for text that happens to line up with them such as bulleted text on a Powerpoint slide and text in general. To eliminate this possibility a second large screen monitor was added so that this type of content can be dragged over to it. The second DVI output drives the 60” LCD display at the right-hand side of the room at 1920×1080 resolution. Windows treats the two monitors as one large virtual display, so content can be easily dragged from the large multi-screen display to the smaller 60” LCD and back.

We wanted the ability to present and control the system from anywhere in the room, so the RF Go Mouse and keyboard were selected. The RF dongle allows us to stay connected from up to 100’ away from the computer, which covers the entire lab and beyond. We tried other wireless keyboards and mice but they quickly lost their connection when they were 10-15 feet away. The 3DConnexion 3D Space Navigator makes it easy to manipulate the Google Earth application, but it is a USB device (no wireless equivalent available yet). To allow us to stretch the Space Navigator anywhere in the room, a USB extender was used to allow us to connect a Cat5 cable as an extension cord for the controller. The same extender was used to allow for placement of the Orbit cam on the opposite side of the room (next to the 60” LCD display).

The Orbit Cam is intriguing as it has a stepper motor in the base which allows the operator to turn it left and right. The auto-focus zoomable lens is able to be moved up and down as well. This allows the operator to pan and zoom anywhere in the room when we’re connected to another researcher or student via Skype or other teleconferencing software.

There is a photo below of me standing next to the multi-display wall with the CEOE website maximized on it. This shows the uber-high resolution of the display and some of the issues that just having it alone (no 2nd display) could cause. The first such monitor that we put in was an 82” Mitsubishi rear-projected LCD display. We ended up returning that display, even though it was larger, because it just wasn’t bright enough. It looked extremely dark when sitting next to the much brighter Samsung LCD display wall.

Samsung-460UX-2-Monitors

Samsung-460UX-2-Monitors

Video Wall Scale

Video Wall Scale

Monitor Wall Mounts

Monitor Wall Mounts

Sharp Aquos 60 inch LCD

Sharp Aquos 60 inch LCD

Logitech Orbit cam

Logitech Orbit cam

USB to Cat5

USB to Cat5

3D Space Navigator

3D Space Navigator

RF Keyboard and Mouse

RF Keyboard and Mouse

Pyle Amp

Pyle Amp

VWBox 133A

VWBox 133A

Dell Precision T7500

Dell Precision T7500

APC UPS

APC UPS

I continue to watch the professional display manufacturer sites for bezel-less LCD displays, which would be my only upgrade that I could imagine for the site. If you run across a 46”+ 1080p zero-bezel display, be sure to send me a link.

The Chief Fusion adjustable wall mounts were quite handy for making minor tweaks to the monitors. It seems that no matter how well you measure, you can never get the displays just perfect, so having the ability to micro-adjust them was quite handy. To allow us to lag-screw them to the wall pretty much anywhere (whether there is a stud or not) we lined the back wall with plywood across the entire wall span and then layered the front with drywall for a finished look. Later on, if we decide to increase the number of monitors into two 9-monitor display arrays, it would be easy enough to add another graphics card, 9 monitors and a second VWBox.

The big secret to turning the project from just a vision to an awe inspiring reality was our most excellent facilities guys and gals. Without their expertise and attention to detail the room could have turned out just ho-hum. They took our ramblings and descriptions of how we’d like things to look and made it come to life. Kudos to them for the room turning out as nice as it did.

Hopefully the information provided here will allow you to build-up your own visualization wall. If you have any questions or comments, please feel free to post them to the site.

Time Lapse Video on the Cheap

The video above is a time lapse of a day in the life of the UD Wind Turbine in Lewes, Delaware.

We were quite excited when they told us that the UD Wind Turbine project was a go. As the time grew near for construction to start, we wanted to chronicle the construction progress and create a time lapse video. I did some research and looked into various webcams with weatherproof housings and the like, but sticker shock at the multi-thousand dollar price tags for the equipment, as well as the networking and power hassles to connect to it made me shy away from a complicated rig. I decided that the best way to go is the simple route.

The task really screamed for a lower cost, battery powered, weather-resistant camera that could be set to take a picture every X number of minutes. I finally narrowed the search down to Wingscapes Birdcam 2.0 outdoor camera. The camera retails for about $200 but I found it on Amazon for just over $150. It has lots of advanced features like motion sensing, light sensing, has a built-in flash plus lots of other nifty features. The main selling points for me were that it was designed for outside use (the turbine was being installed in Spring and it was rainy), it stored its images on an easily accessible secure digital card (up to 4Gigs), it had a user programmable time lapse mode, and it ran on four D-cell batteries for > 4 weeks worth of endurance.

As you can see from the time lapse video that MPEO created of the construction at the turbine base, the results were just what we were looking for (except for the big pile of dirt they put in front of the camera ;?). The video from afar was created using images FTP’d from a webcam located over at the Marine Operations Building. I’ll cover the configuration and components for that webcam setup in a later posting.

I can easily imagine many other uses for this kind device. Time lapse videos of coastal erosion, tide cycles, lab experiment time series, etc. In addition to the features cited above, the camera also has a video and a USB output on the side of the unit as well as an external power connector at the bottom for more lengthy time lapses. All-in-all, highly recommended.

Birdcam Cover Closed

Cover Closed

Birdcam with the cover open

Cover Open

Birdcam Side View

Side View

I used iMovie to create the movie at the top from all of the stills for this post, but I also just as easily created one using the freely downloadable Windows Live Movie Maker if you’re running Windows.

Video Tour of the Research Vessel Hugh R Sharp

RV Hugh R Sharp ready for launchWe recently had guests come down to take a tour of the Lewes campus and the Research Vessel Hugh R Sharp. One of the guests was wheelchair-bound and was limited to only seeing the main deck of the ship as getting to the rest of the ship would have required going up and down stairs. The Sharp has accommodations for handicapped scientists, but they are pretty much limited to the main deck. This limits their access to just the aft working deck, the wet and dry labs, the galley and the conference room. The wheels started turning during that tour on how to share the rest of the technological awesomeness of the Sharp with others. I decided to take my trusty $100 video camera in hand and record a video tour of the ship for those that are unable to navigate the stairs, and for classrooms and visitors who just can’t make the trek to Lewes for a tour. It’s a tad long, running just over 40 minutes or so, but it covers almost the entire ship. Enjoy!

Many thanks to Captain Jimmy Warrington for taking time to do a whirlwind tour just prior to a science mission – as you can tell from the video, he’s a natural at relaying information about the RV Hugh R Sharp and its science capabilities.

Detailed drawings showing deck layouts and profiles of the Sharp can be found the RV Hugh R Sharp landing page, which includes PDFs of:

To help you orient yourself a little bit as to the spaces that were covered, here are some deck diagrams to show the overview of a few of the spaces.

SHARP-AftDeck

Aft Deck

SHARP-DryLab

Dry Lab

SHARP-WetLab

Wet Lab

New Polar and Geosynchronous Satellite Receivers for Delaware

A few weeks ago they fired up a new satellite receiving station from SeaSpace at the University of Delaware’s main campus in Newark, DE. Two receivers were brought online, one for L-Band reception from Geosynchronous Satellites and one for X/L-band reception from Polar Orbiting Satellites. Both receiving systems have dishes that are mounted on the roof of Willard Hall as it presented the least obstructed view of the sky. The adds additional capability to an east coast satellite operations contingent which includes:

  • University of Maine
  • City College of New York
  • Rutgers University
  • University of Delaware
  • University of South Florida
  • Louisiana State University
  • Purdue University

For this blog posting, I’ll only cover the geosynchronous satellite capabilities. In a future posting I’ll cover the polar orbiting hardware and its capabilities.

Geosynchronous Satellite

UD Geosynchronous Satellite Dish

The beauty of geosynchronous satellites is the simplicity with which they can be tracked. Rather than flitting all about and requiring fancy calculations and equipment to track them, you merely point the dish to a point in the sky where the satellite remains fixed relative to the motion of the earth and pretty much lock the receiving dish down. Since the satellite is moving with a trajectory and speed that matches the rotation of the earth, the satellite is said to be “geo-stationary”.

The dish used to receive the signals from the geosynchronous satellites is therefore simple in its design. It is mounted with only one axis of movement, meaning it can only be adjusted along an arc of the sky either to the east or to the west. There is a motor and lead screw mounted on the back that will either push the dish one way, or pull the dish the other in order to position it for the best signal strength. The current intent of the UD dish seems to be dedicated to constantly receiving real-time data from the GOES-EAST satellite (also known as “GOES-13”). GOES East outputs full disk imagery of the the earth from a longitude of 75 degrees west, which gives a good view of pretty much all of North and South America and a good chunk of the Pacific and Atlantic Ocean.

GOES stands for “Geostationary Operational Environmental Satellite” and it is operated by NOAA’s NESDIS or “National Environmental Satellite, Data, and Information Service” primarily to support meteorological operations and research, which includes weather forecasting and storm tracking. The dish is oriented in such a way that it could also be programmed to point to GOES-WEST (aka GOES-11)  for a satellite view of the Pacific Ocean (centered around 135 degrees west longitude) as well if the need arises.

GOES East Full Disk Infrared GOES West Full Disk Infrared

GOES Sensors

One thing to bear in mind is that GOES-13 hasn’t always been “GOES East” – it took over for GOES-12 in April 2010, with GOES-12 moving to 60 degrees West to replace GOES-10 (decommissioned) for coverage of South America. I note this so that you don’t assume that the sensors (and/or their calibration factors)  for a particular GOES station are always the same.

Imager

The current GOES-East has optical imagers with 6 channels with resolutions of 1.1km for the visible channel (one); and 4km and 8km resolutions for the near infrared, water vapor and thermal infrared channels (two through six). The imager is basically a rotating mirror and lens configuration that scans the earth from north to south, line by line to receive reflected visible light, water vapor as well as infrared radiation channels. Each line scanned is digitized and transmitted back towards the earth with measurement units of percent albedo for visible light and temperature for the water vapor and infrared information. Spectral response functions can now also be downloaded online from the NOAA Office of Satellite Operations as well as other GOES calibration information.

Sounder

GOES satellites are also equipped with a sounder with 8km resolution. The sounder scans the atmosphere over the land and ocean and provides vertical profiles which include the temperature of the surface and cloud tops as well as derived wind velocities from these measurements.

Real-time Access to Data

The key feature to having a satellite receiving station on-site is the access to the raw, real-time satellite data. Sure, you can get pull some images down from the NOAA Geostationary Satellite Server, but they would be just derived images. Scientists here at UD and elsewhere are interested in getting the latest raw data feeds from the satellites so that they can research and develop algorithms that process the raw channel data into other products in support of their research projects.

Next on my agenda is to try to give some insight into the polar orbiting satellite tracking station and the fancy gear that sits inside the radome enclosure. Cheers!

REU Intern on FIRe

It has been a great pleasure to have Lauren Wiesebron on the Lewes campus this summer. Lauren is a summer intern from Johns Hopkins and is here as part of an NSF funded Research Education for Undergraduates program (aka REU). For her summer project, Lauren worked in Dr. Matt Oliver’s lab (the ORB Lab) and chose “photosynthetic efficiency” as her summer research project.  To gather data for her project, she set up shop in a portable scientific lab van on the dock in which she set up a Fluorescence Induction and Relaxation System (FIRe Sensor), a Coulter Counter and a Submersible Ultraviolet Nitrate Analyzer (SUNA).  Dr. George Luther’s lab was also taking readings from the same lab van and Lauren included some of their data into her analysis.  The past few weeks Lauren analyzed the results and tomorrow she will present a talk called “Conditions for increased photosynthetic efficiency in an estuarine area”.  Here is a walk-through of the lab van that I did earlier this week with Lauren.

Excellent work Lauren and we hope to see you back here for Grad school!

© 2017 Ocean Bytes Blog

Theme by Anders NorenUp ↑