[18p] REM w/ Ripple

[hhs732]

We are trying to apply the REM idea in Ripple output, then we can use the REM and rating curves idea in HydroVIS, instead of dealing with multiple depth layers.

[hhs732]

The idea of just using the max flow depth layer will not work, because the location of max depth pixel is changing across a reach.

Max depth pixel for flows 275 to 4034 cfs:

Max depth pixel for flows higher than 4655 cfs:

[hhs732]

Algorithem for 2 tiff files of A1 and A2:

Step1: In A1, find the max value (stage) and update A1 cells by subtracting the max from all values above zero (== A1_rem)

Step2: In A2, find the max of A2 and update A2 cells by subtracting all cells above zero from the A2 max, but only do it in cells of A2 that has not been updated in A1. In this step A1 cells updated in step1 will be masked out (== A2_rem).

Step3: in A2, replace the value of rest of A2 cells (not updated in step 2) with A1 cells values. So, only A2 cells that has not be updated in step2 will be replaced (== final_rem).

Step4: Replace zero (0) values in final_rem with the A2 max value (A2 stage) with masking A1_rem

[hhs732]

A1 = np.array([[0,1,0,0,0,0],
               [0,1,2,1,0,1],
               [3,4,6,2,0,2],
               [4,5,8,3,1,3],
               [3,4,6,1,0,1],
               [0,0,0,0,0,0]])

A2 = np.array([[0,2,1,0,0,0],
               [0,2,2,1,1,1],
               [3,4,6,2,2,2],
               [4,6,8,4,3,3],
               [3,5,9,2,0,2],
               [0,0,1,0,0,0]])

# Step 1: Update A1
stage_A1 = np.nanmax(A1)  # Find max value in A1, ignoring NaN
A1_rem = np.where(A1 > 0, stage_A1 - A1, A1)
print(A1_rem)

# Step 2: Update A2
stage_A2 = np.nanmax(A2)  # Find max value in A2, ignoring NaN
mask_A1_rem = (A1_rem != A1)  # Mask of cells updated in A1
print(mask_A1_rem)
A2_rem = np.where((A2 > 0) & ~ mask_A1_rem, stage_A2- A2, A2)
print(A2_rem)

# Step 3: Replace remaining A2 cells with A1 values
mask_A2_rem = (A2_rem != A2)  # Mask of cells updated in A2 during step 2
print(mask_A2_rem)
rem_final = np.where(~mask_A2_rem, A1_rem, A2_rem)
print(rem_final)

# Step 4: Replace remaining o values of A2 cells with stage_A2
ripple_rem = np.where((rem_final == 0) & ~ mask_A1_rem, stage_A2, rem_final)
print(ripple_rem)

rem = [[9 7 8 9 9 9]
[9 7 6 7 8 7]
[5 5 2 6 7 6]
[5 3 0 5 7 5]
[5 4 2 7 9 7]
[9 9 8 9 9 9]]

[hhs732]

Here is a map showing REM for different downstream BC for reach 1291086, HUC8 12030106: