Much like the epic surge that sped forward and burst forth from the banks of the Cedar River on that fateful early June of 2008; the semester for classes has sped by and finals are atop us.
With the clouds now removed from the picture, via the masking discussed in the second post about the final project, the computer can get down to working and showing us visually where the changes occurred between two different satellite images taken by Landsat 5. For our study, I used images captured in June 2007 and June 2008. By comparing the two images in change detection, we can see, literally from space, where the biggest changes (meaning flood water) occurred at.
Presented here in false color imagery, so that vegetation can be easily identified versus that of the rivers and cityscape of Cedar Rapids, is the 2007 and 2008 images. Both have the same cloud mask applied to provide continuity for the change detection.
Seen above via ENVI 5.3 processing, is a change detection difference map. This helps show in a red/green color scale changes the system can detect via powerful algorithms the changes and the value of changes in pixels captured at two different periods in time by Landsat 5. We’re most interested in the blue pixels of the deeper shade, as they represent the largest most profound changes along the banks of the Cedar and Iowa rivers.
Shown above is an alternate processed image that is created via image change workflow toolset in ENVI. This image I find visually provides a more stunning example of the amount of flooding that took place that fateful June. The darkest sections show the most profound changes in the panchromatic imagery. If one looks closely at the city of Cedar Rapids, you can spot the massive damaging floods that took out block after block of homes and businesses in the city. Truly it was an epic surge.
Through our course on Remote Sensing in GIS 255, we’ve been introduced and reminded time after time that radiation is everywhere, except for when there is cloud cover and you’re attempting to utilize a satellite that has atmospheric transmission windows that can be blocked by clouds and heavy water vapor.
Clouds prevent radiation such as sunlight and infrared light from illuminating the area blocked by them or their shadow and that prevention means no information can be reflected back to the sensors on your satellite or other device that is sensing the location.
When you’re limited in respects to where you can get your data from and at what cost you wish to pay, which in this case is nothing, you sometimes do not get the most ideal of results. Above is a false-color image taken by the Landsat 5 satellite on June 16, 2008 which was during the heights of the flooding that had been building up for days prior.
One major obstacle to processing or any analysis was that June 16, 2008 turned out to be a lightly cloudy day which can be seen above in an image that has been edited. The original image had those lovely white fluffy clouds most people enjoy seeing as they laze past during a sunny bright day. Utilizing the ability to capture these clouds and their shadows via the Region of Interest (ROI) tool, I was able to convert the ROI file into a usable mask to hide and remove these features from the imagery, that way when change detection tools would be rolled out, false hits would not be detected from the other image of June 2007 when it was a perfectly clear day without clouds.
While not an ideal or even eye-pleasing end result, it does allow for a more accurate assessment of the size of the flooding that afflicted this portion of the state of Iowa.
Sometimes, just sometimes, seeing is not believing. According to Landsat data collected in 1995 by the Landsat 5 satellite, about 1/5 of the State of Iowa was undergoing a flood of biblical or apocalyptic scales. As you can see, the flood waters stretch an area far larger than the major city of Chicago on the right-hand side of the picture. To cover than much land, which against the popular thought that Iowa is flat, would have been anywhere from dozens to hundreds of feet deep.
Obviously, the real answer to this somewhere an image was mistagged with an incorrect attribute or header on a file and has since lived in infamy since then in the Landsat catalog.
But as a survivor of the Great Imaginary Flood of 1995, I have to give you the cautionary advice: ‘Not everything you can find on the Internet is true’.
For my GIS 295 final project, I decided to connect it to and utilize my work completed on my other course under way in GIS 255. The project is built off of utilizing NASA’s Landsat data to work through a change detection/classification problem.
Originally my idea was based off of viewing the growth of my hometown of Independence, Iowa grow from approximately 1980 up through modern times. The problem that jumped up immediately was an issue of resolution. The town of Independence just wasn’t large enough to easily or accurately view from the scale of the Landsat satellites. So my vision switched to a larger city just south of Independence, that of Cedar Rapids, Iowa.
Once in Cedar Rapids, I came upon an alternative idea of just visualizing growth. In June of 2008, Cedar Rapids and many surrounding towns and cities were inundated by a massive deluge of water that was caused from rivers choked with snow melt and a stalled pressure system that continually drew heavy humidity and moisture north from the Gulf of Mexico.
Change detection to find the areas inundated by this flood was the new goal and task for the projects. But when problem is answered, a new one inevitably will occur. Enter Landsat 7 and its Scan Line Corrector error.
As one can see, the Scan Line Error that manifested back in 2003 is ever present in the 2008 data, making Landsat 7 less than desirable for data collection. Thankfully, Landsat 5 was still operational during the 2008 time period and focus has since switched to utilizing the data collected from that imaging platform to perform analysis.
Anyone who is or knows a farmer or agricultural worker understands the passion and the challenge it is to grow healthy crops. You have so many cards in the deck stacked against you be it the biggest–weather, or a big problem that’s very small–bugs/pests, to the ever invasive Yuppie who bought a house out in the countryside to escape the city but is then mad they have to smell the success of growing food by helping it with manure and other fertilizers.
Technology has shown in the past to be a fantastic hand up on the farmer from the earliest hand-carried tools, to the agricultural revolution that occurred in the 1800s thanks to the ingenious work of John Deere and others in implement equipment. Computer technology is now the new buzz in agriculture on how it can help the farmer grow better, more efficiently, while maintaining a profit without destroying or eroding the land. Data collection is where it lies currently and some of the problems that are slowing it down.
One of the major hurdles that must be tackled aside from the aforementioned weather/pests, would be the act of data. If the data is proprietary and restrictive, there will be little input provided by the main beneficiary but the farmer themselves won’t bother accessing something that could help him/her if it’s cost is beyond any benefit. Open data will be a highly sought after commodity, such as highlighted in this NASA article.
But data is just one part of the larger puzzle of course. Open data will be fantastic but if its without usefulness to the farmer or its unable to be fit into or translated into their needs, it will be useless and a drain upon resources. A big issue with this is the technology versus knowledge gap. If your end-user is 65+ and hasn’t ever used a computer before, giving him an iPad full of programs and some half-baked user manuals is going to see that machine thrown into the corner before the second use. But if you invest in the proper level of education and technology for the end user, it can begin to pay off time invested. Which in the end, should hopefully assist the successful crop bounty.
Palace of Westminster, Courtesy Wikipedia
Everyone has likely heard of the city of London, famous for culture, The Queen of England, tea, football hooligans, and Big Ben. But ever wonder exactly why is London, where it is, instead of say closer to the coast which one would believe when England was and still is a major naval power?
While there has been some prehistoric discoveries of hundreds of years BC showing some minor settlements in the region of London, the biggest reason London is where it is and why it was so successful at remaining is largely due thanks to people laying the way near 2,000 years ago. The Romans.
Rome invaded the British Isles in 43 AD and set forth conquering and securing their new holdings. Their legions filled with tactical and intelligent leaders and engineers saw the advantage of controlling the River Thames at a narrow chokepoint and quickly established a small settlement, known as Londinium on the northern shore of the river. This outpost was short-lived though as in either AD 60 or 61 Boudica led an uprising the burnt the location to the ground but was shortly thereafter dealt with by the Romans.
Undeterred, the Romans resettled the land on the northern banks of the River Thames and laid it out with planned roadways, markets, and most importantly a city wall.
With the fall of the Roman Empire and the subsequent ‘Dark Ages’, Londinium fell into disuse and near abandonment. Eventually as times passed it became used once more thanks in part to the Roman wall and defenses built and the temptation of the location that held the River Thames so well.
Back on November 11th, we were given a guest presentation by one of Fairfax County government’s GIS professionals, Greg Bacon
Greg Bacon, image from his LinkedIn Account
Greg went into an interesting and very detailed discussion about how GIS works for Fairfax County and some of his responsibilities and actions that he does on a day-to-day basis. Undoubtedly a probably very busy man that took time to discuss his passion with a class of individuals looking to break into the business.
What I found most intriguing from the night was when he allowed us to ask questions. I posed him two questions and I’ll paraphrase his answers here as I found them a good assist to the job search dilemma.
Question One: Why do GIS job postings have requirements or basic knowledge requirements of so many different programs, programming languages, etc? Do you need to know all these to get in?
Answer (Paraphrased): The GIS Superman, no I don’t know who writes those job postings but they throw everything they want against the wall and hope for the best. Though you might not know all those requirements, apply anyway, let them be the one’s to ‘weed out’ who is applicable. (In other words, even if you don’t know all the basic requirements they post, apply anyways and let the business run its decision making process and don’t sell yourself short/weed yourself out early because they throw all these terms up and no one is likely to know everything they request.
Question Two: How does one go about building a portfolio to add on top of the school work portfolio when you’re looking for a job?
Answer (Paraphrased): Volunteer work is a great way to build up a portfolio along with internships. Check around with local and county government levels and volunteer your services and time, that way what you’ve done with them is assist them in their end run but also yourself because you’ve provided material to your portfolio.