Table 2 Summary of data included in scenario datasets.

From: Explainable AI approach with original vegetation data classifies spatio-temporal nitrogen in flows from ungauged catchments to the Great Barrier Reef

Dataset type

Purpose

Catchments included

Dataset reference

Variables (discriminated by:)

Target data

LU spatial variables

OV spatial variables

EU spatial variables

90 day prior flows

Corresponding DIN

Corresponding catchment

1. Spatial Classification Training Datasets

Training algorithms to match spatial data to corresponding gauged catchment

All gauged

1. GCAFALU

✓

–

–

–

–

✓

1.GCAFAOV

–

✓

–

–

–

✓

1.GCAFAEU

–

–

✓

–

–

✓

1.GCAFAEULUOV

✓

✓

✓

–

–

✓

2. Spatial Classification Testing Datasets

Classify pseudo-ungauged and ungauged Catchments to gauged catchment

Individual ungauged/pseudo-ungauged

2.CAFALU

✓

–

–

–

–

Hidden

2.CAFAOV

–

✓

–

–

–

Hidden

2.CAFAEU

–

–

✓

–

–

Hidden

2.CAFAEULUOV

✓

✓

✓

–

–

Hidden

3. ANN-WQ simulator development- training datasets

Establish whether recognisable patterns exist in the datasets to forecast DIN

All gauged together (non-discriminated)

3.GCAFALU

✓

–

–

✓

✓

–

3.GCAFAOV

–

✓

–

✓

✓

–

3.GCAFAEU

–

–

✓

✓

✓

–

3.GCAFAEULUOV

✓

✓

✓

✓

✓

–

All gauged—individual (Discriminated by catchment)

3.G\(\text {C}_{{\text{i}_1}...{\text{i}_{\text n}}}\)FALU

✓

–

–

✓

✓

–

3.G\(\text {C}_{{\text{i}_1}...{\text{i}_{\text n}}}\) FAOV

–

✓

–

✓

✓

–

3.G\(\text {C}_{{\text{i}_1}...{\text{i}_{\text n}}}\) FAEU

–

–

✓

✓

✓

–

3.G\(\text {C}_{{\text{i}_1}...{\text{i}_{\text n}}}\) FAEULUOV

✓

✓

✓

✓

✓

–

Gauged Catchments grouped by Spatio-Temporal Category (Discriminated by Category 1, Category 2 or Category 3)

3.GC1…3F1…3LU

✓

–

–

✓

✓

–

3.GC1…3F1…3OV

–

✓

–

✓

✓

–

3.GC1…3F1…3EU

–

–

✓

✓

✓

–

3.GC1…3F1…3EULUOV

✓

✓

✓

✓

✓

–

4. ANN-WQ Simulator-Trial Datasets

Evaluate suitability of matched dataset for data transfer to pseudo-ungauged catchment for DIN simulation purposes

Individual gauged and pseudo-ungauged catchment

4.\(\text {C}_{{\text{i}_1}...{\text{i}_{\text n}}}\)FALU

✓

–

–

✓

Hidden

–

4.\(\text {C}_{{\text{i}_1}...{\text{i}_{\text n}}}\) FAOV

–

✓

–

✓

Hidden

–

4.\(\text {C}_{{\text{i}_1}...{\text{i}_{\text n}}}\) FAEU

–

–

✓

✓

Hidden

–

4.\(\text {C}_{{\text{i}_1}...{\text{i}_{\text n}}}\) FAEULUOV

✓

✓

✓

✓

Hidden

–

Gauged and pseudo-ungauged catchments grouped by categories

4.C1…3F1…3LU

–

✓

–

✓

Hidden

–

4.C1…3F1…3OV

–

✓

–

✓

Hidden

–

4.C1…3F1…3EU

–

✓

–

✓

Hidden

–

4.GC1…3F1…3EULUOV

✓

✓

✓

✓

Hidden

 
  1. Variations in each dataset intended to evaluate the influence of spatial data or flow variable toward DIN response.
  2. G = Gauged, C Catchment, F Flows, A All, LU LandUse, OV Original Vegetation, EU Ecounits \(C_{i_1...i_n}\) reference to individual catchments.
Back to article page