Had meeting with the Remote Sensing expert from Oceanography and my advisor. We talked over what I had learned so far in the quarter and got some interesting new ideas about how to proceed.
The most interesting, and possibly viable, one was the use of Spectral Unmixing. This means taking the value of a pixel and then divide it up by its components since no pixel is really homogeneous, especially at a 30m resolution. I think this will work, if I can learn to do it.
I'm reading some articles on it right now. Lots of them deal with fire, which was where I had originally learned about it, but there are studies using this technique for almost any subject that uses Remote Sensing. I need to read a bit more to understand the particulars but if I can learn to use the program, and not calculate anything myself, I think this would be very useful.
Wednesday, May 18, 2011
Wednesday, May 11, 2011
Day 8--ECHO and other things
So after following some good advice first my adviser to e-mail someone, and then from that person's response to e-mail someone else, I have some confirmed information on how to obtain free ASTER data. I just need to sign up for an ECHO account which, is not as easy as it sounds. I attempted to follow the link I was given (https://api.echo.nasa.gov/pump/faces/register_pump.jsp) I could not actually sign up (the page was not available). I e-mailed back to inquire and will probably try again later.
It also seems possible that we will be able to obtain the data since we will only need a couple of bands and a few images, only the visible red and near infrared to calculate NDVI.
Also, I wanted to explore again the price of Landsat-5 Data and it still appears to be free. There is a quote on the main USGS website saying "The opening of the Landsat archive to free, web-based access is like giving a library card for the world's best library of Earth conditions to everyone in the world". So I assume I can download it for free over the internet.
It also seems possible that we will be able to obtain the data since we will only need a couple of bands and a few images, only the visible red and near infrared to calculate NDVI.
Also, I wanted to explore again the price of Landsat-5 Data and it still appears to be free. There is a quote on the main USGS website saying "The opening of the Landsat archive to free, web-based access is like giving a library card for the world's best library of Earth conditions to everyone in the world". So I assume I can download it for free over the internet.
Thursday, April 28, 2011
Day 7-MODIS Details
Vegetation Indices Products 16 and Monthly time scales & 250m-5600m resolution
The 500m 16-day Vegetation Indices product provides consistency for usage and centers blue at 469nm, red at 645nm, and NIR to 858nm which are used to defined daily vegetation indices. It tries to maintain sensitivity over dense areas (which we are interested in) and uses the EVI(Enhanced Vegetation Index) to correct for atmospheric contamination by smoke and sub-pixel thin clouds. The product is computed from a corrected product. The data is aimed for use in monitoring land cover changes as well as inputs for modeling climate change.
The lower resolution MOD13A2 is the same sort of thing but at 1000m instead. The 250m band since it lacks a 250m blue band it uses the 500m to correct for atmospheric effects. The image of this one looks less crisp than the 500m one. The 1000m monthly image uses a weighted average of the the data if it is cloud free.
Burned Area Monthly 500m
Uses daily surface reflectance values to look for rapid changes indicative of burning and maps the extent of recent fires only. It gives a quality "score" per pixel. It is based on 3 months of corrected daily reflectance data
Thermal Anomalies and fire at 1000m and 5 min, Daily, and 8 day time scales
It derives these areas from micrometer radiances based on its difference from background and is used to advance the monitoring of fires globally. The Daily one has three dimensions, the fire-mask, fire radiative power, and are given for an 8-day period. The 8 day product gives the average of these days which gives a fire-mask and the algorithm quality.
Leaf Area Index 8 day time scale 1000m resolution
It defines the number of equivalent layers of leaves per unit ground area and are used to calculated surface photosynthesis etc.
Land Cover Type Yearly Time scale and 1000m and 5600m
Taking a years worth of data it classifies the areas into 17 different classes defined by the International geosphere Biosphere Programme.
Opinion
The one 250m product seems too derived to work well but the multiple 500m look useful. However, these are much bigger than we really want to use. The 1000m are useful if we want to look at more 'general' trends that is not the focus. I think the combined burned, a vegetation index, and a leaf area index would be most useful. Preferably at 500m and looking at the daily to 16-day cycles rather than the monthly.
The 500m 16-day Vegetation Indices product provides consistency for usage and centers blue at 469nm, red at 645nm, and NIR to 858nm which are used to defined daily vegetation indices. It tries to maintain sensitivity over dense areas (which we are interested in) and uses the EVI(Enhanced Vegetation Index) to correct for atmospheric contamination by smoke and sub-pixel thin clouds. The product is computed from a corrected product. The data is aimed for use in monitoring land cover changes as well as inputs for modeling climate change.
The lower resolution MOD13A2 is the same sort of thing but at 1000m instead. The 250m band since it lacks a 250m blue band it uses the 500m to correct for atmospheric effects. The image of this one looks less crisp than the 500m one. The 1000m monthly image uses a weighted average of the the data if it is cloud free.
Burned Area Monthly 500m
Uses daily surface reflectance values to look for rapid changes indicative of burning and maps the extent of recent fires only. It gives a quality "score" per pixel. It is based on 3 months of corrected daily reflectance data
Thermal Anomalies and fire at 1000m and 5 min, Daily, and 8 day time scales
It derives these areas from micrometer radiances based on its difference from background and is used to advance the monitoring of fires globally. The Daily one has three dimensions, the fire-mask, fire radiative power, and are given for an 8-day period. The 8 day product gives the average of these days which gives a fire-mask and the algorithm quality.
Leaf Area Index 8 day time scale 1000m resolution
It defines the number of equivalent layers of leaves per unit ground area and are used to calculated surface photosynthesis etc.
Land Cover Type Yearly Time scale and 1000m and 5600m
Taking a years worth of data it classifies the areas into 17 different classes defined by the International geosphere Biosphere Programme.
Opinion
The one 250m product seems too derived to work well but the multiple 500m look useful. However, these are much bigger than we really want to use. The 1000m are useful if we want to look at more 'general' trends that is not the focus. I think the combined burned, a vegetation index, and a leaf area index would be most useful. Preferably at 500m and looking at the daily to 16-day cycles rather than the monthly.
Day 7-Readings
So yesterday I was able to catch up on some reading mostly on remote sensing projects in the Mediterranean region. Each article had a different focus and use different data sources, two of which were not sensors we had thought to use in this project. These are summaries for only two of them.
Estimating spectral separability of satellite derived parameters for burned areas mapping in the Calabria region by using SPOT-Vegetation data 2006
The study area for this article was Southern Italy and compared known burned areas to their image on remotely sensed data. They chose to do this because of the affect of wildfires in the Mediterranean region since a small fire can in fact have a big impact. Their overarching purpose was to see if remote sensing could help to evaluate the disturbance by fire of and testing fire models. They used 10-image composites from June till September 1998 received from the Vlaamse Instelling voor Technologisch Ondersock (VITO) Image Processing center which has free vegetation products. They then compared this to the Italian National Forestry Services record of fires for that time period. They eventually came up with different indices depending on the area within their study zone. They suggest that a better exploration would be to discrimination of areas depending on the land cover type, such as pasture v. forests, and that their processes could be applied using different sensors, such as MODIS-Terra.
An integrated spatial and spectral approach to the classification of Mediterranean land cover types: the SSC Method 2004
This explores a new way of classifying remotely sensed data by taking into account the idea that undefined pixels are most likely to be closely related to those nearby. It wants to use this principle to help define "open" types of land cover that do not have definite boundaries, such as shrub vegetation or vineyards. They used ENVI to do their project and relied on three main steps in their method.
1) Stratification: which was used to find "homogeneous" regions based on spatial and spectral data.
2) Classification of those homogeneous regions
3) Classification of the rest of the "heterogeneous" image
They used a lot of equations to defined exactly how "similar" mixed pixels were and whether or not they could be incorporated into a closer homogeneous area. Using these principles and equations they compared their remotely sensed classification to ground proofing classifications. They tested two regions, one most open land and the other mostly farmland. They found that this method classified open area 8% better than the regular method but that "closed" areas did not have an improved classification. In their acknowledgements they state that their methods are available on request which might be useful for the what we are doing. However, its in ENVI which I don't know how to use.
Estimating spectral separability of satellite derived parameters for burned areas mapping in the Calabria region by using SPOT-Vegetation data 2006
The study area for this article was Southern Italy and compared known burned areas to their image on remotely sensed data. They chose to do this because of the affect of wildfires in the Mediterranean region since a small fire can in fact have a big impact. Their overarching purpose was to see if remote sensing could help to evaluate the disturbance by fire of and testing fire models. They used 10-image composites from June till September 1998 received from the Vlaamse Instelling voor Technologisch Ondersock (VITO) Image Processing center which has free vegetation products. They then compared this to the Italian National Forestry Services record of fires for that time period. They eventually came up with different indices depending on the area within their study zone. They suggest that a better exploration would be to discrimination of areas depending on the land cover type, such as pasture v. forests, and that their processes could be applied using different sensors, such as MODIS-Terra.
An integrated spatial and spectral approach to the classification of Mediterranean land cover types: the SSC Method 2004
This explores a new way of classifying remotely sensed data by taking into account the idea that undefined pixels are most likely to be closely related to those nearby. It wants to use this principle to help define "open" types of land cover that do not have definite boundaries, such as shrub vegetation or vineyards. They used ENVI to do their project and relied on three main steps in their method.
1) Stratification: which was used to find "homogeneous" regions based on spatial and spectral data.
2) Classification of those homogeneous regions
3) Classification of the rest of the "heterogeneous" image
They used a lot of equations to defined exactly how "similar" mixed pixels were and whether or not they could be incorporated into a closer homogeneous area. Using these principles and equations they compared their remotely sensed classification to ground proofing classifications. They tested two regions, one most open land and the other mostly farmland. They found that this method classified open area 8% better than the regular method but that "closed" areas did not have an improved classification. In their acknowledgements they state that their methods are available on request which might be useful for the what we are doing. However, its in ENVI which I don't know how to use.
Monday, April 25, 2011
Day 6-Landsat 5 and Landsat 7
So the Landsat System is a collection of satellites that have provided continuous information about the earth's surface. There have been 7 launches one of which (landsat 6) failed. The ones we want to focus on in this project are Landsat 5 and Landsat 7, although the strict focus is on landsat 5 since landsat 7 has some malfunctions causing spots on the images. They follow a ground track in a 185km swath that goes from north to south and around the same time every 16 or 18 days it passes over the same spot on the earth.
Landsat 5 combines a Multispectral scanner (MSS) with a Thematic Mapper (TM) to have four spectral bands from visible green to near-infrared and a shortwave infrared and an improve resolution of 120m with the thermal-IR band and of 30m on the rest of the available 6 bands. The .pdf at http://pubs.usgs.gov/fs/2010/3026/pdf/FS2010-3026.pdf also explains what each band means and what it can be used for. It also gives a pretty good description of landsat 7.
The malfunction I mentioned is described in this document as a "scan line corrector failure". I don't think its that great. There is apparently a way to "fix" it using the data but looking at figure 6 I don't think it is really as accurate as they want it to be.
Landsat 5 combines a Multispectral scanner (MSS) with a Thematic Mapper (TM) to have four spectral bands from visible green to near-infrared and a shortwave infrared and an improve resolution of 120m with the thermal-IR band and of 30m on the rest of the available 6 bands. The .pdf at http://pubs.usgs.gov/fs/2010/3026/pdf/FS2010-3026.pdf also explains what each band means and what it can be used for. It also gives a pretty good description of landsat 7.
The malfunction I mentioned is described in this document as a "scan line corrector failure". I don't think its that great. There is apparently a way to "fix" it using the data but looking at figure 6 I don't think it is really as accurate as they want it to be.
Day 6-Landsat
Much better website and the two interfaces to look for data are much better. However, the question or accessibility to cost eludes me. From the information in Sarah Parcak's book it could range from nothing to $600. The search tool on the Landsat website seems to make no reference to cost and on at least one page (http://pubs.usgs.gov/fs/2010/3026/pdf/FS2010-3026.pdf) there is a disclaimer stating "All USGS Landsat data acquired from 1972 to the present are available over the Internet at no charge and with no user restrictions." So it seems possible.
Accessing Data
The two different interfaces one can use to access the data are much better than those for MODIS and ASTER, though I think you can use them and I really should. Glovis (http://glovis.usgs.gov/)and EarthExplorer (http://edcsns17.cr.usgs.gov/NewEarthExplorer/) are graphical so you can look at the images you want to download. Although there is sort of odd coverage of places. I also think I need to register to download stuff directly. But much better than ASTER!
Accessing Data
The two different interfaces one can use to access the data are much better than those for MODIS and ASTER, though I think you can use them and I really should. Glovis (http://glovis.usgs.gov/)and EarthExplorer (http://edcsns17.cr.usgs.gov/NewEarthExplorer/) are graphical so you can look at the images you want to download. Although there is sort of odd coverage of places. I also think I need to register to download stuff directly. But much better than ASTER!
Day 6-ASTER, Cost
After poking around on a very, very bad website I finally found a costs table for the different data types available from ERSDAC website. Data levels 1-2 are 9,800 yen per scene which, at the current exchange rate, is about $120. Not too much, but truly not great, if we wanted to use more than one image that starts getting to be more than I'd like to spend. Also I am not quite sure if this is accurate anymore since it hasn't been updated sine March 2004 (at least this section of the webpage). I would assume then that it has gone up in the recent years but I can't know until I put in an order, which I am reluctant to do at the moment.
Though they do give a good flow chart as to what you should do with the webpage, found at http://www.gds.aster.ersdac.or.jp/gds_www2002/exhibition_e/a_products_e/set_a_produ_e.html.
On to Landsat-5
Though they do give a good flow chart as to what you should do with the webpage, found at http://www.gds.aster.ersdac.or.jp/gds_www2002/exhibition_e/a_products_e/set_a_produ_e.html.
On to Landsat-5
Subscribe to:
Posts (Atom)