Dirty REMOTE SENSING Lecture 3: First Steps in classifying Stuart Green Earthobservation.wordpress.com

Transcription

1 Dirty REMOTE SENSING Lecture 3: First Steps in classifying Stuart Green Earthobservation.wordpress.com

2 You have your image, but is it any good? Is it full of cloud? Is it the right time of year? Is it the right resolution?

3 Atmospheric effects The atmosphere, as we have seen, can interfere with the picture the satellite is taking. Clouds and their shadows are obvious but atmospheric effects that are not readily visible can distort the image, reduce the signal/noise ratio, introduce errors in classification

4 Atmospheric Correction Raw Image: The image recorded at the satellite- this is uncalibrated data that contains all the errors from the atmosphere and the sensor. Level 0 Level 1 usually corrects for satellite level errors and presents the data in Digital Numbers. Sometime terrain corrections are included.

5 Top of Atmosphere reflectance correction converts the L1 DN to a number representing the percentage of light that fell on the earth as reflected back to sensor. Full atmospheric correction, to surface reflectance. This is either a supplied commercial image or using free software like LEDAPS you can perform this correction yourself.

6 Atmospheric correction corrects for scattering and absorption in the atmosphere. At its most sophisticated it used models of the atmosphere like 6S along with estimate of optical parameters of the atmosphere from satellites. At its simplest- we use the dark target approach

7 If you need accurate, transferable, repeatable measurements you will need to atmospherically correct your imagery. This will mean your image values are equivalent to surface measurements. HOWEVER Landsat data are available as a atmospherically corrected data set (level 2a)

8 First Automation find Vegation Sources Online for Vegetation detection :net_primary_productivity ftp://ftp.biosfera.dea.ufv.br/users/francisca/franciz/papers/running %20et%20al.%20Bioscience% pdf AES2012 L6 NDVI

9 Video: Global Seasonal Vegetation Growth: a planetary phenology AES2012 L6 NDVI

10 How do we measure what veg is there? Biomass: The mass per unit area of vegetation. Cover: The vertical projection of the plant parts on the ground surface per unit area of ground. Usually expressed as a percent. No species can have more than 100% cover. Leaf Area Index: The ratio of the area of leaves and green vegetation in theplant canopy per unit area of ground surface. LAI can exceed 1.The only way to get true leaf area is to strip all the leaves off the plants and measure their area. All other methods provide an index of this value Normalized Difference Vegetation Index (NDVI): An index of vegetation greenness derived from remote sensing methods. Often used as an index of biomass. FPAR measures the proportion of available radiation in the photosynthetically active wavelengths (400 to 700 nm) that a canopy absorbs. AES2012 L6 NDVI

11 AES2012 L6 NDVI

12 AES2012 L6 NDVI

13 AES2012 L6 NDVI

14 We can measure with the sateelite sensors the ratio of red to NIR light- this ratio is called a Vegetaion Index AES2012 L6 NDVI

15 Vegetation Indices? The gigantic chlorophyll absorption well distinguishes vegetation from nonvegetation. Its size tells us chlorophyll concentration in the leaf and the canopy. Many vegetation indices are a simplistic attempt to estimate the size of this absorption well. AES2012 L6 NDVI

16 Vegetation Indices Vegetation indices (VI s) can be broken up into two basic categories: Ratio based indices VI s based on the ratio of two or more radiance, reflectance, or DN values (or linear combinations thereof). Difference indices VI s based on the difference between the spectral response of vegetation and the soil background. AES2012 L6 NDVI

17 Common Ratio Indices Simple Ratio Index (SR) = NIR/R Normalized Difference Vegetation Index (NDVI) = NIR NIR R R AES2012 L6 NDVI

18 What are Vegetation Indices? Estimating the size of the absorption well 0.5 reflectance(%) density 1 density 2 density 3 density 4 density 5 density 6 sunlit soil wavelength AES2012 L6 NDVI

19 (B4-B3)/(B4+B3) for Landsat A pixel by pixel mathematical process AES2012 L6 NDVI

20 AES2012 L6 NDVI

21 Remember the images are stored in the image file (eg. Jpeg) as xy matrices with a pixel in one band corresponding with the pixel in another band with same xy coordinates B B2 AES2012 L6 NDVI B1+B2

22 Other Indices SAVI = NDII= (VNIR-NIR)/(VNIR+NIR) AES2012 L6 NDVI

23 Create and NDVI image See And

24 Practical

25 Load Curragh17.Tiff into ARC MAP and open the ToolBox window

26 WE need to convert the image to a floating point format

27 Do this for Band 4 (red) and Band 5 (NIR). Call the outputs RED and NIR

28 Use Map Calculator to calculate the NDVI image

29 This is the calculation Press OK

30 Use the Info button to click around and see the values

31 Load up the Sites.shp file What are the NDVI values for the 6 sites. What s the relationship between landcover and NDVI? (list the landcover types and NDVI score) Click around the image- find an NDVI value that represents no vegeation

32 Click onto reclass->reclassify and load your NDVI image into pop up menu- we are going to create a vegeation/no-vegetaion map.

33 On the reclassify menu click classify and change the number of classes to 2. Then change the 1st break values to be your bare soil ndvi value (o.3 in this example) and the second to 1. Click OK

34 Clcik the output button and enter a value a name in YOUR directory do not add an extension. Click OK

35 Your output will look a little like this. You have created a vegetation/no-vegetation mask