Preparing Qext for Sandbox (commit 8c1479b)

[c-h-david]

Dev Note: Preparing Qext for Sandbox

Motivation

RAPID requires a sandbox for controlled verification and testing.
This note documents the steps used to create that sandbox.
More information is at SANDBOX.md

The associated dataset is published on Zenodo:

This corresponds to commit 8c1479b.

---

Design Considerations

• The signal should remain constant within each day but vary between days.
  → Each of the eight 3-hourly time steps per day contains identical data, producing a square-wave pattern.

• Parameter k = 9000 s; a day (86,400 s) is roughly 9.6× longer, so the system should reach steady state daily.

• The initial discharge value should equal the final discharge value.

• The same program should be able to create both the true and flawed Qext datasets.

• Bias corrections should:

  • Differ between sub-basins
  • Remain consistent within each sub-basin
    This enables reconstruction of the “truth” using both data assimilation (DA) and bias correction.

---

Steps

1. Create Qext for Sandbox

sandboxqext -m 10.0 10.0 10.0 20.0 20.0 \
            -a 1.0 1.0 1.0 2.0 2.0 \
            -o input/Sandbox/Qext_Sandbox_19700101_19700110.nc4

---

2. Create Initial State (all zeros)

zeroqinit -i input/Sandbox/Qext_Sandbox_19700101_19700110.nc4 \
          -o input/Sandbox/Qinit_Sandbox_19700101_19700110.nc4

---

3. Set Initial State Values

import netCDF4
e = netCDF4.Dataset('input/Sandbox/Qinit_Sandbox_19700101_19700110.nc4', 'a')
e.variables['Qout'][0, :] = [9, 9, 27, 18, 63]
e.close()

---

4. Run RAPID Simulation

rapid2 -nl input/Sandbox/namelist_Sandbox.yml

---

5. Rename Output Files

mv output/Sandbox/Qout_Sandbox_19700101_19700110_tst.nc4   output/Sandbox/Qout_Sandbox_19700101_19700110.nc4
mv output/Sandbox/Qfinal_Sandbox_19700101_19700110_tst.nc4 output/Sandbox/Qfinal_Sandbox_19700101_19700110.nc4

---

6. Export Example CSVs (for reach 10)

import netCDF4
import csv

# Input Qext
f = netCDF4.Dataset('input/Sandbox/Qext_Sandbox_19700101_19700110.nc4', 'r')
ZV_Qex = f.variables['Qext'][:, 0]
IV_tim = f.variables['time'][:]
f.close()

# Output Qout
g = netCDF4.Dataset('output/Sandbox/Qout_Sandbox_19700101_19700110.nc4', 'r')
ZV_Qou = g.variables['Qout'][:, 0]
g.close()

# Write CSVs
with open('Qex.csv', 'w', newline='') as csvfile:
    writer = csv.writer(csvfile)
    writer.writerow(['time', 'Qext'])
    writer.writerows(zip(IV_tim, ZV_Qex))

with open('Qou.csv', 'w', newline='') as csvfile:
    writer = csv.writer(csvfile)
    writer.writerow(['time', 'Qout'])
    writer.writerows(zip(IV_tim, ZV_Qou))