2. NDVI Anomaly, Kenya, January 2009

3. Vegetation Indices Enhancing green vegetation using mathematical equations and transformations

4. Learning Objectives What are vegetation indices? What do we hope to accomplish with them? Understand the relationship between spectral indices and spectral reflectance curves. What features of vegetation spectra are most indices based on? What are advantages and disadvantages of various algabraic indices?

5. What is a “vegetation index”? A mathematical combination or transformation of spectral bands that accentuates the spectral properties of green plants so that they appear distinct from other image features.

6. What Should Vegetation Indices Do?? Indicate the AMOUNT of vegetation (e.g., %cover, LAI, biomass, etc.) Distinguish between soil and vegetation Be insensitive to atmospheric and topographic effects if possible

7. How is Vegetation Spectrally Distinct? Reflectance in individual wavelength regions (bands)? Shape of spectral curve created by looking at more than one wavelength region? Changes in spectral curves with amount of vegetation? Others?

8. Soil Reflectance Can be bright in NIR (like vegetation) –dry soil especially bright –wet soil much darker than dry soil Soil can have low visible light reflectance (like vegetation)

9. Group Exercise Given typical green vegetation spectral reflectance, and reflectance of soils ranging from dark to bright, propose an algebraic combination of two Landsat 8 bands that will distinguish the plants from the soils!

10. Vegetation vs. Soil and Water

11. How can we use this with digital imagery? Many vegetation indices are based on accentuating the DIFFERENCE between red and NIR reflectance in image pixels Big Difference Small Difference

12. Difference Vegetation Index (DVI) Probably the simplest vegetation index –Sensitive to the amount of vegetation –Distinguishes between soil and vegetation –Does NOT deal well with the difference between reflectance and radiance caused by the atmosphere or shadows So for example…can’t distinguish vegetation from soil in shady areas very well. A problem when there is topography.

13. Ratio-based Vegetation Indices Simplest ratio-based index is called the Simple Ratio (SR) or Ratio Vegetation Index (RVI) –High for vegetation –Low for soil, ice, water, etc. –Indicates the amount of vegetation –Reduces the effects of atmosphere and topography

14. Why Simple Ratios Reduce Atmospheric and Topographic Effects

15. Problem with SR Division by zero Wide range of possible values depending on amount of red reflectance These problems addressed by development of the NDVI

16. Normalized Difference Vegetation Index NDVI = (NIR – Red)/(NIR + Red) –Ranges from -1 to 1 –Never (Rarely?) divide by zero –Indicates amount of vegetation, distinguishes veg from soil, minimizes topographic effects, etc. –A good index! –Does not eliminate atmospheric effects!

17. NDVI Applications

19. But…Problem with NDVI (and some other ratios) Sensitive to soil background reflectance Non-linear changes in index as amount of vegetation changes Not insensitive to atmosphere Affected by geometry Saturation problems So…use with caution. Great for many applications but not all!

20. Soil Background Effects

21. IR G R B (Amount changes depending on soil)

22. Indices get “tuned” to try to reduce these problems. E.g., Soil Adjusted Vegetation Index (SAVI) –Uses a soil background “fudge factor” SAVI = [(NIR – Red)/(NIR + Red + L)] * (1 + L) L is a soil fudge factor that varies from 0 to 1 depending on the soil. Often set to 1.

23. Vegetation Amount (LAI)

25. Choosing an Algabraic Index Most difference indices fall short in terms of dealing with atmospheric and topographic effects Most ratio-based indices are functionally equivalent (work about the same) Some ratio-based indices are computationally “cleaner” NDVI is often the index of choice and generally performs pretty well, but you must be aware of potential issues

26. Next Lecture… Indices based on data transformations and “feature space”