We have moved! Please visit us at ANTHROECOLOGY.ORG. This website is for archival purposes only.


Mar 28 2012

Surveying update

Over spring break I created eight rough points in with the Field GPS. These points are scattered around the knoll and Herbert Run. I placed four points new points around The Knoll that all look into the forest one on Admin drive two on the loop and one near the parking garage. I beleive with these points I will be able to survey the entire Knoll.

I placed four extra survey points scattered through out Herbert Run that will most likely be able to finish the survey there. One of the Points is on the hill . One near the swing and two more scattered in strategic locations the obtain line of sight to the rest of the points.

Today I called Will Wiley and he was in a meeting. My plan is to go out to Herbert run and The Knoll with him and Have him make RTK GPS points for the rough points I created.

As soon as Will calls we will go out and create these points. Then I will survey plot fifty first thing. After that I Will finish the primarly grid in HR. Once that is done I will Immediately do The Knolll. I expect the time it takes to finish HR run to be very very quick because I  laid down so many control points that no matter what I must have a good line of sight some where this way I can rapidly move from one point to the next and never spend any time trying to squeeze a point out of a tight survey location.

The Knoll will also be pretty quick for all the peripheral points. There is a chance that I Will need one extra point that is at the top of The KNoll where the elevation levels off. This area might be hard to get from the four points I have created. I will bring this up with Will and see if he can give me a ninth point on top of thi hill.

 At the bottom are all the new points I have located to be new control points.

Mar 28 2012

Forest Fire Videos Captured with a Hexakopter

Here are links to two videos we collected of the forest fire we flew in early March in New Jersery.  Stephen had posted previously, post here, about our experiences getting the payload to fly on the Hexakopter, a repeat of similar work from last year.

Taking video was an afterthought for the project so we only had an SD4000 on hand for filming.  I think the video quality would have been a lot better if we had the GoPro.  Despite the wiggle and bluriness of the video, it is still possible to see the smoke lines set by the fire managers (see images at left) and there are places in the afternoon video where I think you can even see fire.  Note that these videos were taken pointing about due west, so I believe the shots from later in the day appear darker due to exposure compensation.

Video from about 12:45 - 12:55 local time: http://youtu.be/q3S3V3212bo?hd=1

Video from about 15:45 - 15:55 local time: http://youtu.be/GQHh2LX_ad8?hd=1

Mar 27 2012

Topography and the Mapping Grid

There has been a new data sheet designed to address the specific needs of the forest we are working with. Because the method for mapping the trees has changed, the data sheets also needed to be altered. We are returning to the previous used method of laying out a 1x1 meter grid within our 5x5 meter grid. Once this is complete the location of the trees will be marked on the graph found on the data sheet. There has also been a "codes" column added to the data sheet to represent trees that may need special attention. This could include a leaning stem, a stem broken below breast hight, or as seen in the picture multiple stems from one trunk forming below breast height. However, before the trees can be mapped the grid must first be sectioned into 5x5 meter squares. Jonathan, fellow students, and I are hoping to get one of the 25x25 meter plots sectioned off so we can begin to test our tree mapping stratgies. We are also tackling the problems we may face concerning drastic elevation changes. In summary we have all of our supplies ready and in bags we just need to find a time to get dirty and see how our ideas work.


 

Mar 21 2012

Tree mapping Technique

There have been many methods for mapping the trees within our 25x25 meter grid that we have identified. The one certatinty we have decided on is the grid must be sectioned into a 5x5 meter grid before we can begin mapping. The picture on the left shows a method found in the field guide Methods For Establishment And Inventory Of Permanent Plots. This method involves usining geometry to determine the exact point of a tree and we thought it could be more accurate and faster than other ideas. However when we went to our forrest to test we discovered that it was not only more tedious but may not improve accuracy by a reasonable amount if at all. The problems arose when we needed to take measurments on unlevel surfaces. It would involve 3 or more people with much instruction and using handfulls of equpment, it was uneffective for our purposes. We plan on going on another test run before the week ends to try another method that will hopefully work for what we need. 

References:

Dallmeier, F. (1992). "Long-term monitoring of biological diversity in tropical forest areas." Methods for establishment and inventory of permanent plots. MAB Digest Series, 11. UNESCO. Paris

Mar 21 2012

Herbert Run Update

On Monday (3/21/2012), Andrew and I went to Herbert Run to survey more points to make a grid so that we can start mapping trees.  By the end of the day, we finished enough points to have ten 25 by 25 grid points marked and ready for tree mapping.  We begin mapping trees today (3/21/2012).

Mar 20 2012

New Jersey Pinelands Fire Research Flight

This past Tuesday (3/6/12) Jonathan and I travelled to New Jersey to conduct wildfire research.

We left Monday evening and drove to the New Jersey Pinelands, where we stayed overnight at the research center.  Our first obstacle was the smoke research payload.  When Jonathan did this smoke research in the past, he had suspended the payload (consisting chiefly of a smoke detector and data logger) on a tether beneath the hexakopter.  The reasoning behind this was that we needed to keep the payload well away from the air disturbance made by the hexa's propellers.  However, the problem with this method is that the hexakopter is unable to self stabilize under a suspended payload, and ends up swinging the payload wildly and eventually crashing.

Jonathan and I did a brief test to confirm that this behavior was still present with the fire payload, it was.  As a solution, we mounted the payload on top of the hexakopter's dome.  A strip of ribbon attached to the top of the payload served to demonstrate that the propellers did not significantly interfere with the airflow though the payload when it was mounted on top.  The hexakopter was perfectly stable with the top mounted payload. 

Tuesday during the wildfire we flew from the research compound which was downwind of the wildfire in the smoke plume.  The hexakopter flew up in 50 meter increments and back down every half hour or so for several hours.  Overall our data collection was a success.  Towards the end we even had the time to mount a camera on top of the payload to take a rather jittery video of the wildfire from the air.

Mar 16 2012

Updates: Hexakopters, Quad Arducopter and Octocopter.

Now that the hexakopters are basically complete (one has a programming error that we are currently working on), Stephen and I decided to work with our other projects: quad arducopter and octocopter.

The quad arducopter (4 blades) is working.  We spent a few days familiarizing ourselves with the parts of the copter using the arducopter wiki (http://code.google.com/p/arducopter/wiki/ArduCopter).  The electronic speed controller (ECS) still needs to be calibrated, but it the copter is connecting to the computer and seems to be in working order.  We have started the parts list so that we can move forward.

The octocopter (8 blades) is still in the supply obtaining phase.  We have already placed our order for the necessary parts to construct the copter.  Within a few weeks we shall have those supplies and can commence the construction!  The objective of the octocopter is to run the trials the hexakopters are currently doing, but more efficiently.  It is also hoped that we can use the octocopters to expand the scanning area. 

Mar 07 2012

Finishing the survey grid

There have been many issues that have presented themselves while trying to survey this particular section of woods. There are many drastic elevation changes and two of our points appear to lie in the middle of the only small stream found in this section of forest. The surveying must be completed before the trees start to bud otherwise visibility would be virtually impossible to obtain, this being chief issue with the completion of the survey grid. When there is an accurate survey point set we can only plot a half a dozen points before needing a new point to shoot from due to visibility issues. Initially we discovered a way to make a new reference point through resection however Andrew and I soon discovered this was not an accurate enough approach to finishing the grid. A few more reference points must be measured before the grid can be completed and tree marking begins. An accurate effective method of tree marking must be used and we are in the process of obtaining Dr. Richard Condit's Tropical Forest Census Plots: Methods and Results from Barro Colorado Island, Panama and a Comparison with Other Plots. Hopefully this will give us a new outlook on how to plot the trees in the most precise manner.

Mar 05 2012

Welcome New Ecosynth Users!

Just a quick post welcoming Jessica Breen and Steve Klosterman as new members of the Ecosynth blog.

Jessica will be pursuing a Ph.D. with Dr. Ellis in the UMBC GES Department this coming fall.  She has had experience with kite aerial photography and is interested in citizen science remote sensing.  She keeps up a blog about her experiences, An Accidental Geographer.

Steve is pursuing a Ph.D. at Harvard University with Dr. Andew Richarson.  He is interested in deploying multirotors for personal remote sensing of forest canopy phenology.

Welcome Jessica and Steve!

Mar 04 2012

Traversing vs. resection

  After careful research I realized my understanding of resectioning was not clear. Resectioning a control point is not the most efficient way to establish new control in alot of ways its a waste of time. A better way of survey control setting would be traversing.

With resection I noticed that you can use 3 known points to determine a unknown point.The problem with this is, I would need to be set up some where roughly in the middle of the three points because according to the owners manuel for the total station, the further I am from the center of the points the less accuracy I will have in creating my new control points. I observed this to be true because no matter how many new points I created they were all off by and angle of at least 1'.

This led me to consider another method of creating control; traversing(picture shown below). Traversing is the method that all modern surveying is done with. This method allows me to broaden my survey network while also creating new control points. Triangualation deals with an X,Y cooridinate system and the pythagorean theorum in order to determine the exact differences between many points. Traversing uses a R,Theta coordinate system of vectors to determine the relative distances between many points. Where you would use a magnitude R and a direction determined by some angle "Theta". What you see below is a "closed traverse". This is the most widely used surveying method today because it is quick and efficient. basically what you do is you begin and end on the same point calling them different points. You have at least three traverse in a loop. What you do is collect many measurements of the distances and angles inbetween three points. By setting up on one point and backsighting and forsighting two other points. After this it will propogate the error and evenly distrubute it throughout the three points. This process is repeated howevermany times until you have a sufficient amount of control points to cover all your mapping and stake-out needs. When your loop reaches the first point agian, you would simply give that point the next highest number and throw a control point there. Then after this you would do a closure check which has two components. First I would check the relative distances between the two points. For example, in the figure below if I were to call the point that the total station is one traverse 106 and where the forsight is the next point counter clockwise ( you always throw traverse counter clockwise, this is shown but not said in the figure) traverse 107/101 then I could collect the sets and check realively speeking how close 107 is to 101. The ideal situation is that they have the same coordinates exactly which is never the case. After the closure process. You can use the data collectore to basically distribute all the error in this ring throughout the ring and make a super accurate set of control points. Finally I would use the equation (n-2)180 and if the angle is withing (+-) some relatively small angle difference, then the traverse loop has good closure. For example on this one below we have six angles

(6-2)180degrees=720degrees 

Therefore if the angle was 720degrees00'05" then I would definitely keep it because this traverse has good closure.

In order to do this I will need a data collector. Here is an advertisement that shows the whole process of traversing and also could be a good data collector to buy

Traversing by Repetition With the RECON Data Collector - YouTube

I think we need to get a data collector as well as the other equiptment that I have put on the Surveying equipment order.