← Back to Index
Chapter 9

River Network Generation

Syllabus & Marks Distribution

DetailInformation
Chapter9 — River Network Generation
Teaching Hours4 hours
Marks Allocation6 marks
Total Course Marks80 marks
Weight7.5% of total

Syllabus Topics

  • Flow direction
  • Flow accumulation
  • River network
  • Watershed boundary delineation

Importance & Frequently Asked Topics HIGH

Marks Distribution by Each Exam

Asked 17 times across past exams. Total marks: 161. Average marks per exam: ~10.

ExamMarks
2071 Bhadra12
2071 Magh6
2072 Ashwin8
2072 Magh6
2073 Bhadra12
2073 Magh10
2074 Bhadra16
2074 Magh15
2075 Bhadra6
2075 Magh14
2076 Bhadra8
2078 Chaitra8
2079 Chaitra10
2080 Chaitra12
2081 Chaitra10
2082 Shrawan8

Frequently Asked Topics

TopicTimes AskedFrequency
Flow Direction 16
Flow Accumulation 1

Topic-wise Importance & Exam Tips

TopicImportanceKey Points / Exam Tips
Flow direction (D8 algorithm)HighNumerical problems very common. Given a DEM grid, calculate flow direction for each cell using steepest descent. Know the 8-direction encoding (1,2,4,8,16,32,64,128).
Flow accumulation calculationHighFrom flow direction grid, compute how many cells drain into each cell. Grid-based numerical problems are exam staples.
River network extractionHighApply threshold to flow accumulation. Explain how threshold value affects network density. Stream ordering (Strahler method).
Watershed boundary delineationHighTrace upstream from pour point using flow direction. Draw watershed boundary on grid. Extremely important for exam.
DEM preprocessing (sink filling)MediumWhy sinks must be filled before hydrological analysis. Depressionless DEM concept.
Complete numerical problem (DEM → flow direction → accumulation → network → watershed)HighFull chain problems carrying 6+ marks are common. Practice multiple grid examples.

1. The Foundation: Watersheds and the DEM

PAST EXAM QUESTIONS ON THIS TOPIC:

  1. Define watershed and explain filled DEM and flow direction [1+5] (2072 Magh, Q10)

Definition of a Watershed:

A watershed (also known as a catchment area or drainage basin) is a topographically defined geographical area of land where all surface water (from rain, melting snow, or ice) converges and drains off into a single, common outlet point, usually known as a "pour point" (like a river mouth, a lake, or a dam). In GIS, the boundary of a watershed is defined by topographic ridges (drainage divides) extracted from a Digital Elevation Model (DEM).

Watershed (Catchment Area) Concept Rain Surface runoff Surface runoff Drainage Divide Drainage Divide Pour Point (Outlet) Stream Legend Ridge Stream Outlet Basin All precipitation within the watershed boundary drains to the single pour point
Fig 9.1 – Watershed concept: ridges define drainage divides, all surface water converges to a single pour point

2. Pre-processing the DEM: Sinks and Filled DEMs

PAST EXAM QUESTIONS ON THIS TOPIC:

  1. Define watershed and explain filled DEM and flow direction [1+5] (2072 Magh, Q10)
  2. Explain how a sink is developed in DEM and how it can be removed [4] (2074 Bhadra, Q2a)
  3. Explain sinks and filled DEM (2075 Magh, Q3a)

Before any hydrological analysis can begin the raw DEM must be checked for errors.

What is a Sink?

A sink (or depression) is a cell or a group of adjacent cells in a DEM that are completely surrounded by cells with higher elevation values. Because water flows downhill by gravity any water that enters a sink is trapped and cannot flow out causing the digital river network to artificially disconnect and stop.

How is a Sink Developed in a DEM?

  1. Data Errors (Artifacts): Most sinks are artificial errors created during the data collection process (e.g. radar shadow in SRTM or ASTER data) or during the spatial interpolation of the DEM grid.
  2. Resolution Limitations: If the grid resolution is too coarse (e.g. 90m pixels) a narrow gorge might be missed making it look like a closed basin.
  3. Natural Sinks: Occasionally they are real geographical features such as glacial hollows quarries or karst (limestone) sinkholes.
    4. Three Causes of Sinks in DEMs 1. Data Errors (Artifacts) Radar signal Shadow SRTM / ASTER radar misses some areas Most common cause 2. Resolution Too Coarse 90 85 92 88 90m pixel ? 30m pixel Narrow gorge missed by coarse grid Solution: use finer DEM 3. Natural (Real) Sinks Lake Sinkhole Glacial hollows quarries karst/limestone sinkholes Do NOT fill these!
    Fig 9.2 – Three causes of sinks: data artifacts (most common) coarse resolution limitations and real natural depressions

How is it Removed? (The "Fill" Operation)

Sinks are removed using the Fill Sink geoprocessing tool. The software algorithm identifies the trapped cells and artificially raises their elevation values until they are perfectly level with the lowest surrounding boundary cell (the "pour point" of the sink).

[X] Before: DEM with Sink trapped ⬇ Sink Water cannot drain out → Fill [OK] After: Filled DEM ⬆ Raised to pour point flow → Continuous drainage restored
Fig 9.3 – Sink and Fill Operation: Raising trapped depressions to allow continuous flow

What is a Filled DEM?

A Filled DEM (or Depressionless DEM) is the final output of the Fill operation. It is a corrected elevation surface where all artificial sinks have been eliminated ensuring that water can flow continuously across the landscape to the edges of the map.

3. Core Hydrological Algorithms

PAST EXAM QUESTIONS ON THIS TOPIC:

  1. Explain Sink, Flow Direction, Flow Accumulation, River Grid, and stream lines [5] (2074 Magh, Q5a)

Once you have a Filled DEM you run a sequence of specific tools to create the rivers.

A. Flow Direction

Flow Direction calculates the exact path water will take as it moves from one pixel to the next. The standard method used is the D8 (Eight-Direction) Algorithm. The software evaluates the center cell and its 8 immediate neighbors. It calculates the drop in elevation and assigns a flow direction code pointing to the single neighboring cell that has the steepest downward slope.

D8 Direction Codes NW32 N64 NE128 W16 X E1 SW8 S4 SE2 Powers of 2 encode 8 directions Example: Steepest Drop → SE 78 72 69 74 67 56 68 58 46 Result Direction = 2 (Southeast) 67→46 = −21m Largest drop determines flow code
Fig 9.4 – D8 Flow Direction Algorithm: Direction codes and example (67→46 is steepest drop)

B. Flow Accumulation

Using the flow direction network the Flow Accumulation tool calculates how much water flows into every single cell. It does this by counting the total number of upstream cells that pour into a specific downslope cell.

Flow Accumulation: Counting Upstream Cells 78 72 69 71 58 74 67 56 49 46 69 53 44 37 38 64 58 55 22 31 Acc=3 Acc=5 How it works: Each cell counts how many upstream cells drain into it. Cells with high accumulation values = stream channels. Apply threshold: if Accumulation > T classify as stream. Lower T = denser network. Green highlighted cells show higher accumulation (convergence points = likely streams).
Fig 9.5 – Flow Accumulation: each cell counts upstream contributing cells; high values indicate stream channels

C. River Grid (Stream Definition)

Not every cell with flow accumulation is a river; some are just muddy slopes. To define an actual river you apply a Threshold using Map Algebra. For example you might set a rule: If Flow Accumulation > 500 cells then it is a River (1) else it is Not a River (0). The output is a binary River Grid where the pixels representing streams are highlighted.

Stream Threshold Effect on River Density Low Threshold (T = 100) Dense network Medium Threshold (T = 500) Moderate network High Threshold (T = 2000) Only main river Rule: If Flow Accumulation > T cell is classified as stream. Lower T = denser network.
Fig 9.6 – Effect of threshold value on stream density: low thresholds capture all tributaries high thresholds show only major rivers

D. Stream Lines (Stream to Feature)

The River Grid is still a raster (pixels). To make it useful for mapping it is converted into vector polylines using the "Stream to Feature" tool. These resulting vector lines are the Stream Lines.

Stream to Feature: Raster -> Vector Conversion Binary River Grid 0 0 1 0 0 0 0 1 1 1 1 1 Pixels (1 = river 0 = land) Stream to Feature Vector Polylines (Shapefile) Smooth connected polylines Attribute Table ID | Length | Order 1 | 2.4 km | 2 Raster pixels become smooth vector lines with attributes (length order connectivity)
Fig 9.7 – Stream to Feature conversion: binary raster river grid is converted into vector polylines for mapping and analysis

4. Stream Ordering: The Shreve Method

PAST EXAM QUESTIONS ON THIS TOPIC:

  1. Describe about the Shreve method for stream order (2075 Bhadra, Q9b)
  2. What are the steps involved in generating a river network within a given area (represented by a polygon vector data) assuming that multiple DEMs are available which covers the whole area? Explain steps with sketches. (2082 Shrawan)
  3. What are the steps involved in generating a river network within a given area (represented by a polygon vector data) assuming that multiple DEMs are available which covers the whole area? Explain steps with sketches. (2082 Shrawan)

Once stream lines are created hydrologists classify their size and importance using stream ordering. While the Strahler method is the most common the exams specifically ask for the Shreve Method.

The Shreve Method:

In the Shreve system all exterior links (the smallest starting streams with no tributaries) are assigned a magnitude of 1.

Unlike other methods when two streams intersect their magnitudes are strictly added together.

Shreve Method: Strictly Additive 1 1 1 1 1 2 1+1 2 1+1 3 2+1 5 3+2 Outlet (total magnitude = 5 tributaries)
Fig 9.8 – Shreve Stream Ordering: Magnitudes are strictly additive at confluences
Stream Ordering: Strahler vs Shreve Methods Strahler Method Order increases only when same-order streams meet 1 1 2 1 2 Rule: 1+1=2 2+1=2 2+2=3 Shreve Method Magnitudes always add at every confluence 1 1 2 1 3 Rule: 1+1=2 2+1=3 2+2=4 (always additive) VS
Fig 9.9 – Strahler ordering increases only when equal orders meet; Shreve always adds magnitudes at confluences

5. The Complete Workflow / Scenario Question

This is a comprehensive 10-mark question that requires you to outline the entire process including initial data management.

Question Workflow: Generating a River Network from Multiple DEMs

Step 1: Data Preparation (Mosaic)

Since multiple DEMs cover the area the first step is to use the Mosaic to New Raster tool. This stitches the multiple separate DEM grids together into one large continuous regional DEM.

Step 2: Area Extraction (Clip/Extract by Mask)

Use the provided "Polygon Vector Data" (the study area) as a mask. Use the Extract by Mask (or Clip) tool to cut the large mosaicked DEM down to the exact boundary of your study area to save processing time.

Data Preparation: Mosaic + Clip for Multi-DEM Projects Raw DEM Tiles Tile A Tile B Tile C Tile D 4 separate rasters Mosaic Merged DEM One continuous raster Study area polygon Seamless surface Clip Clipped DEM Only study area retained Ready for Fill -> Flow Dir Steps 1-2 of the workflow: combine multiple DEMs then extract the study region before hydrological processing
Fig 9.10 – Data preparation steps: Mosaic merges multiple DEM tiles then Clip (Extract by Mask) trims to the study area boundary

Step 3: Remove Depressions (Fill)

Run the Fill tool on the clipped DEM to remove any sinks and create a depressionless DEM.

Step 4: Calculate Flow Direction

Input the Filled DEM into the Flow Direction tool. This applies the D8 algorithm to determine the downslope path for every pixel.

Step 5: Calculate Flow Accumulation

Input the Flow Direction raster into the Flow Accumulation tool. This calculates the total number of upstream cells contributing to each pixel.

Step 6: Stream Definition (Map Algebra/Raster Calculator)

Apply a threshold to define the rivers. Use the Raster Calculator with a conditional statement: Con("Flow_Accumulation" > Threshold_Value 1). This outputs a binary River Grid.

Step 7: Convert to Vector (Stream to Feature)

Finally use the Stream to Feature tool (inputting both the binary River Grid and the Flow Direction raster). This converts the raster pixels into smooth vector polylines resulting in your final continuous River Network shapefile.

Complete River Network Generation Workflow (7 Steps) 1. Raw DEMs Multiple tiles from source -> 2. Mosaic Merge + Clip to study area -> 3. Fill Sinks Remove false depressions -> 4. Flow Dir D8 algorithm 8 directions -> 5. Flow Acc Count upstream contributing cells -> 6. Threshold Acc > T -> stream Binary river grid -> 7. Stream to Feature -> Vector Raster Raster Raster Raster Raster Raster Vector! Key Decision Points: • Step 3 is critical: unfilled sinks block flow and create false flat areas (broken drainage) • Step 6 threshold value controls stream density: low T = many streams high T = only major rivers • Entire workflow is raster-based until final Step 7 converts to vector polylines for mapping
Fig 9.11 – Complete 7-step river network generation workflow from raw DEMs to vector stream network

Solved Numerical 1

PAST EXAM QUESTIONS ON THIS TOPIC:

  1. Find the sink cell; Show flow direction; Calculate flow accumulation; Sketch rivers; Find area of biggest watershed [2+2+2+3+3] (2080 Chaitra, Q10)

Question: Following diagram shows an elevation raster data at 30 m spatial resolution. Refer to any cell as columns (A to F) and rows (1 to 6). For example the value of cell B2 is 118. Answer the questions based on this elevation raster.

Given: Elevation Raster (30 m resolution) A B C D E F 1 2 3 4 5 6 130 128 102 101 124 138 135 118 97 88 114 121 110 98 84 68 92 103 115 86 81 78 76 87 112 102 84 73 96 105 129 118 103 85 119 105
Fig 9.12 – Given elevation raster (30 m). Highlighted cells D3=68 and D5=73 are sinks.

(a) Find the Sink Cells and Fill

PAST EXAM QUESTIONS ON THIS TOPIC:

[2 Marks]

𝑓 sink_fill.calc
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
// ── Sink Detection & Filling · 30m DEM ──
// A sink = cell lower than ALL 8 neighbors
Sink 1: D3 = 68
Neighbors: C2=97 D2=88 E2=114 C3=84 E3=92 C4=81 D4=78 E4=76
Minimum neighbor = 76 (E4)
68 < 76 → SINK [OK]
Fill: 76 + 0.5 = 76.5
Sink 2: D5 = 73
Neighbors: C4=81 D4=78 E4=76 C5=84 E5=96 C6=103 D6=85 E6=119
Minimum neighbor = 76 (E4)
73 < 76 → SINK [OK]
Fill: 76 + 0.5 = 76.5
// After filling: D3 = 76.5 D5 = 76.5
// E4 (76) now becomes lowest interior cell → drainage outlet
// All 36 cells drain toward E4
Hydrology · DEMSink Fill · 18 lines

(b) Flow Direction for All Cells [2 Marks]

Using the D8 algorithm on the filled DEM, each cell flows to the neighbor with the steepest downhill gradient (drop/distance; diagonal distance = 30√2 ≈ 42.43 m).

D8 Flow Direction (Arrows) OUT All 36 cells drain to E4 (outlet)
Fig 9.13 – D8 Flow Direction arrows. All cells converge toward E4 (lowest point = drainage outlet).
𝑓 flow_direction_Q1.calc
1
2
3
4
5
6
7
8
9
10
11
// ── Flow Direction Summary (D8 Algorithm) ──
// A B C D E F
Row 1: SE E SE S SW S
Row 2: SE SE SE S SW SW
Row 3: SE E E SE S SW
Row 4: E E E E OUT W
Row 5: NE E E E N NW
Row 6: NE NE N N W NW
// Pattern: North side flows S/SE, South side flows N/NE
// Everything converges toward E4 (drainage outlet)
Hydrology · D8Flow Direction · 11 lines

(c) Flow Accumulation [2 Marks]

Flow accumulation counts how many upstream cells drain through each cell. A cell with FA=0 is a ridgeline; high FA values indicate channels.

𝑓 flow_accumulation_Q1.calc
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
// ── Flow Accumulation Grid ──
// A B C D E F
Row 1: 0 0 1 0 0 0
Row 2: 0 1 0 4 0 1
Row 3: 0 1 4 12 2 0
Row 4: 0 3 4 5 35 0
Row 5: 0 1 4 7 1 0
Row 6: 0 0 0 1 0 0
// ── Key drainage paths ──
// North channel: D2(4) → D3(12) → E4(35)
// West channel: B4(3) → C4(4) → D4(5) → E4(35)
// South channel: C5(4) → D5(7) → E4(35) (via D5→E4)
// Verification: E4 = 35 upstream cells
// 35 + 1 (self) = 36 = 6×6 total cells [OK]
How to count:
Start from ridge cells (FA=0), trace downstream.
Each cell's FA = sum of (upstream cell's FA + 1) for all cells flowing into it.
Example: D3 receives C2(0+1) + C3(4+1) + D2(4+1) + E2(0+1) = 1+5+5+1 = 12
Hydrology · FAFlow Accumulation · 23 lines

(d) River Network (threshold = 4500 m²)

PAST EXAM QUESTIONS ON THIS TOPIC:

[3 Marks]

𝑓 river_network_Q1.calc
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
// ── Stream Definition ──
Given: Cell size = 30 m, Threshold = 4500 m²
Cell area = 30 × 30 = 900 m²
Min FA = 4500 / 900 = 5 cells
// Cells with FA ≥ 5 → RIVER
River cells:
D3 = 12 [OK]
D4 = 5 [OK]
D5 = 7 [OK]
E4 = 35 [OK]
// River sketch:
// D3 flows SE to E4 (north tributary)
// D4 flows E to E4 (west tributary)
// D5 flows E to E4 (south tributary)
// Three branches converge at E4 (confluence)
Hydrology · RiverStream Definition · 17 lines
River Network (FA ≥ 5) 001000 010401 01420 0340 01410 000100 12 5 35 7
Fig 9.14 – River network: three tributaries (D3, D4, D5) converge at E4 (confluence point)

(e) Area of the Biggest Watershed

PAST EXAM QUESTIONS ON THIS TOPIC:

[3 Marks]

𝑓 watershed_Q1.calc
1
2
3
4
5
6
7
8
9
// ── Watershed Area ──
Finding: ALL 36 cells drain to E4 (single outlet)
→ Entire grid is ONE watershed
Area:
Cells = 6 × 6 = 36
Cell area = 30 × 30 = 900 m²
Total = 36 × 900 = 32,400 m² = 3.24 ha
Hydrology · WatershedArea Calc · 9 lines

Solved Numerical 2

PAST EXAM QUESTIONS ON THIS TOPIC:

  1. The following grid represents a DEM data of certain area whose cell size is 100 m. Using this data answer the following questions (a) Show the sink cell (b) What will be the value of that cell when the sink is filled? (c) Calculate flow direction and flow accumulation value for each cell (d) What is the area of the largest possible watershed? (e) Show the river in the raster when stream definition is limited to minimum catchment area of 40000 m▓ [10] (2079 Chaitra, Q8)

Question: The following grid represents a DEM data of certain area whose cell size is 100 m. Using this data answer the following questions: (a) Show the sink cell. (b) What will be the value of that cell when the sink is filled? (c) Calculate flow direction and flow accumulation value for each cell. (d) What is the area of the largest possible watershed? (e) Show the river when stream definition is limited to minimum catchment area of 40000 m².

Given: Elevation Raster (100 m resolution) A B C D E F 1 2 3 4 5 6 78 72 69 71 58 49 74 67 56 49 46 50 69 40 44 37 37 48 64 58 55 22 22 24 78 61 47 21 21 19 74 61 47 11 11 12
Fig 9.15 – Given elevation raster (100 m). Highlighted cell B3=40 is the only sink.

(a) Sink Cell & (b) Fill Value

PAST EXAM QUESTIONS ON THIS TOPIC:

[2 Marks]

𝑓 sink_fill_Q2.calc
1
2
3
4
5
6
7
8
9
10
11
12
// ── Sink Detection · 100m DEM ──
Sink: B3 = 40
Neighbors: A2=74, B2=67, C2=56, A3=69, C3=44, A4=64, B4=58, C4=55
Minimum neighbor = 44 (C3)
40 < 44 → SINK [OK]
Fill value:
Pour point = lowest rim cell = C3 (44)
B3 (filled) = 44
// After filling, B3 = 44. Water at B3 now drains east toward C3.
Hydrology · DEMSink Fill · 12 lines

(c) Flow Direction & Flow Accumulation

PAST EXAM QUESTIONS ON THIS TOPIC:

[4 Marks]

𝑓 flow_dir_Q2.calc
1
2
3
4
5
6
7
8
9
10
11
// ── Flow Direction (D8, steepest descent) ──
// A B C D E F
Row 1: SE SE SE S S SW
Row 2: SE S SE S S SW
Row 3: E E SE S S S
Row 4: NE N E S SE S
Row 5: E E E S S S
Row 6: E E E OUT OUT W
// D6 and E6 = outlets (lowest boundary cells, elev 11)
// Two watersheds: western (→D6) and eastern (→E6)
Hydrology · D8Flow Direction · 11 lines
𝑓 flow_accum_Q2.calc
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
// ── Flow Accumulation Grid ──
// A B C D E F
Row 1: 0 0 0 0 0 0
Row 2: 0 1 1 2 2 0
Row 3: 0 6 7 5 4 0
Row 4: 0 0 0 15 5 1
Row 5: 0 1 2 19 0 8
Row 6: 0 1 2 23 11 9
// ── How to compute (example) ──
B3 (FA=6):
Receives: A2(0+1) + A3(0+1) + A4(0+1) + B2(1+1) + B4(0+1) = 6
D4 (FA=15):
Receives: C3(7+1) + C4(0+1) + D3(5+1) = 8+1+6 = 15
D6 (FA=23):
Receives: C6(2+1) + D5(19+1) = 3+20 = 23
// Verification:
// D6 watershed: 23+1 = 24 cells
// E6 watershed: 11+1 = 12 cells
// Total: 24+12 = 36 = 6×6 [OK]
Hydrology · FAFlow Accumulation · 24 lines

(d) Area of the Largest Watershed

PAST EXAM QUESTIONS ON THIS TOPIC:

[2 Marks]

𝑓 watershed_Q2.calc
1
2
3
4
5
6
7
8
9
10
11
12
13
14
// ── Two Watersheds (two outlets at row 6) ──
Watershed 1 (outlet = D6)
Cells: A1-A6, B1-B6, C1-C6, D1-D6 = 24 cells
Area = 24 × (100 × 100) = 240,000 m² = 24 ha
Watershed 2 (outlet = E6)
Cells: E1-E6, F1-F6 = 12 cells
Area = 12 × 10,000 = 120,000 m²
Answer:
Largest watershed = Watershed 1 (D6)
Area = 240,000 m² = 24 ha
Hydrology · WatershedArea Comparison · 14 lines

(e) River Network (min catchment = 40,000 m²)

PAST EXAM QUESTIONS ON THIS TOPIC:

[2 Marks]

𝑓 river_Q2.calc
1
2
3
4
5
6
7
8
9
// ── River Network (min catchment = 40,000 m²) ──
Given: Cell Area = 100 × 100 = 10,000 m²
Threshold: 40,000 m²
Min FA = 40,000 / 10,000 = 4 cells
River Cells (FA ≥ 4):
C4(4), C5(5), D3(5), D4(10), D5(15), D6(23)
C6(6), B5(4), B6(5), E5(4), E6(11)
Hydrology · RiversRiver Network · 9 lines

Interactive Hydrology Solver

Use this interactive tool to explore the complete hydrology workflow step-by-step. Enter your own 6×6 DEM values and resolution, then click through each tab to see sink detection, flow direction, flow accumulation, river network, and watershed delineation computed in real time.

📊 Enter DEM Data

Hover over any cell on the grid to inspect data
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
// ── Stream Definition ──
Given: Cell size = 100 m, Threshold = 40,000 m²
Cell area = 100 × 100 = 10,000 m²
Min FA = 40,000 / 10,000 = 4 cells
// Cells with FA ≥ 4 → RIVER
Western river (Watershed 1 → D6):
B3 = 6 [OK]
C3 = 7 [OK]
D3 = 5 [OK]
D4 = 15 [OK]
D5 = 19 [OK]
D6 = 23 [OK]
Eastern river (Watershed 2 → E6):
E3 = 4 [OK]
E4 = 5 [OK]
F5 = 8 [OK]
F6 = 9 [OK]
E6 = 11 [OK]
// Total river cells: 11 out of 36
Hydrology · RiverStream Network · 22 lines
River Network (FA ≥ 4) 000000 011220 00 0001 0120 012 6 7 5 15 19 23 4 5 8 9 11 Western river (→D6, 24 ha) Eastern river (→E6, 12 ha)
Fig 9.16 – River network: Western channel (B3→C3→D3→D4→D5→D6) and Eastern channel (E3→E4→F5→F6→E6)

Chapter Summary

This chapter covers the GIS-based process of extracting river networks and watershed boundaries from Digital Elevation Models (DEMs). The process follows a systematic sequence: first, flow direction is determined using the D8 (deterministic eight-neighbor) algorithm, which assigns each cell a flow direction toward the steepest downslope neighbor. Next, flow accumulation counts the number of upstream cells draining through each cell — cells with high accumulation values represent stream channels. A threshold value is applied to the flow accumulation grid to extract the river network (cells above the threshold are classified as streams). Finally, watershed boundary delineation identifies the drainage area contributing flow to a specified outlet point by tracing flow directions upstream. The chapter also covers DEM preprocessing steps like filling sinks (depressionless DEM), stream ordering (Strahler and Shreve methods), and pour point identification. These techniques are fundamental for hydrological modeling, flood risk analysis, and water resource management.

End of Chapter 9 – River Network Generation