E Exercise 03

Last updated: 2021-03-31 00:29:17

E.1 Question 1

• Read the raster MOD13A3_2000_2019.tif to a multi-band raster object named r.
• Calculate a vector of length 24 with the correlation coefficients between the values of the layer 1, and the values of layers 1, 2, …, 24.
• To calculate correlation coefficients use the cor function with the use="pairwise.complete.obs" argument, as in: cor(x, y, use="pairwise.complete.obs").
• For example, the 2^nd value in the vector should be equal to:

## [1] 0.9752619

• Plot the values of the vector, as shown in Figure E.1.
• Remember: you cannot use the raster package in your solution! (see Section B.5.5.)
• Note: the two arguments passed to the cor function, i.e., the pixel values of layer 1 and layer n, need to be either vectors or arrays (not matrices). If necessary, a matrix can be converted to a vector with as.vector.

Figure E.1: Correlation between NDVI values in layer 1 and layers 1-24

(50 points)

E.2 Question 2

• Read the raster MOD13A3_2000_2019.tif to a multi-band raster object named r.
• Create a single-band raster s, marking areas with relatively high NDVI, relatively low NDVI, and large amount of NA, as follows:
  • Pixels where the maximal non-na NDVI value across all 233 bands of r is ≥0.2, will have a value of 1 in s
  • Pixels where the maximal non-na NDVI value across all 233 bands of r is <0.2, will have a value of 0 in s
  • Pixels where the proportion of NA values is >0.2 (regardless of NDVI values), will have a value of NA in s
• Plot the raster s with values of 0 marked in "orange" and values of 1 marked in "darkgreen" (values of NA will be transparent, by default)
• Remember: you cannot use the raster package in your solution! (see Section B.5.5.)

Figure E.2: \(NDVI_{max}≥0.2\) (green) and \(NDVI_{max}<0.2\) (orange)

(50 points)