# Site02 | Flooding, Water-Level Change and Ecology

English edition v1, expanded observation atlas and flood history · Prepared and retrieved: 20 September 2026 · Intended use: landscape research, site analysis and concept discussion.

**Website update:** The explorer now includes a detailed, filterable flora/fauna observation map, a four-event historical peak comparison, and the documented 2010 gauge milestones. See [the atlas and history guide](History_and_Observation_Atlas_Guide.md) for controls, source pages and interpretation limits.

**Unified website:** Water-level scenarios, ecological observations, flood history and both future time scales now share [one research atlas](Site02_Water_and_Ecology_Explorer.html). Use **Future** in its navigation; the previous future-page link redirects to the same atlas.

## How to interpret this package

Site02 is on the Queanbeyan River near Morisset Street and Buttles Creek. The boundary was copied from the Site02 feature referenced by the supplied APRX. Its area is **68.206 ha**; it is not a newly estimated boundary. The original APRX and the Chinese edition remain unchanged.

This is an **inland freshwater river reach**. The relevant processes are rising and falling river levels associated with rainfall, upstream inflows and reservoir regulation, rather than coastal tides. Sea-level, intertidal and mangrove models should not be applied to this site. See [BOM flood types](https://www.bom.gov.au/resources/learn-and-explore/flood-knowledge-centre/about-floods) and [BOM Canberra water-resource context](https://www.bom.gov.au/water/nwa/2015/canberra/contextual/climateandwateroverview.shtml).

The package separates three kinds of evidence:

1. **Official material:** the 2020 floodplain study, NSW extant and pre-clearing vegetation models, the river corridor strategy and the 2024 platypus/rakali report.
2. **Occurrence records:** GBIF observations with dates, reported coordinates, uncertainty and source links.
3. **Calculated scenarios:** DEM-based connected static stage sensitivity. These scenarios are uncalibrated and **are not a two-dimensional hydraulic model or flood forecast**.

Use official flood studies as the primary basis for flood-risk interpretation. Present the generated scenarios as terrain-sensitivity and water–land relationship studies. They do not replace the official model.

## 1. Open and use the deliverables

- `Site02_Water_and_Ecology_Explorer.html`: offline English viewer with a slider for 17 offsets from 0 to 8 m and a rise/fall sequence. Map images are embedded; no online basemap is required.
- `Site02_Flood_Ecology.aprx`: four independent maps. Toggle `Scenario_02m`, `04m`, `06m` and `08m`; +6 m is the main default scenario.
- `GIS/Flood_Ecology.gdb`: native vector data. `GIS/GeoJSON`: WGS84 interchange files. `GIS/*.tif`: georeferenced 1 m rasters.
- `Maps/01_Flood_Scenarios.png`: four-stage comparison.
- `Maps/02_Vegetation_History.png`: extant versus reconstructed pre-clearing communities.
- `Maps/03_Ecology_Observations.png`: fauna and flora observations aggregated to a 500 m grid.
- `Maps/04_Model_Reference.png`: the reference profile and area response.
- `Maps/05_Rise_Fall_Schematic.gif`: presentation-ready animation of the static sequence.
- Matching SVGs retain editable text and vector annotations. Terrain and depth components remain embedded rasters.

This package supplements the earlier hydrology and river-history collection. Use that earlier collection for the wider river network, upstream reservoirs, historic aerial-photo index and catchment context. The English edition changes presentation and documentation only; numerical tables, GIS geometry and model values are retained.

## 2. How the water-level scenarios were calculated

### Inputs and extent

| Item | Input used |
|---|---|
| Horizontal CRS | GDA94 / MGA zone 55, EPSG:28355; metres |
| Terrain | Supplied ACT2015 DEM, 1 m cells; not a new 2026 survey |
| Vertical datum | Unverified. Values are source DEM Z and must not be labelled AHD |
| Boundary | Original Site02 feature, 1018 × 670 m |
| Model frame | Rectangle extending approximately 500 m beyond the site, limited by available DEM coverage |
| Valid terrain coverage | 100% inside the site; 86.03% across the model frame |
| Connectivity seeds | NSW mapped water polygon, `C_WaterAreas` |
| Probability and time | No AEP, return period, real duration or rate of stage change assigned |

### Method

Within the mapped water polygon, the 20th percentile of DEM values is calculated in 20 m northing bands. A five-band median filter smooths these samples, followed by a non-decreasing southward fit. The profile is interpolated to every raster row and extended laterally to create `Reference_Surface_Not_Observed.tif`.

This is a **constructed reference**, not an observed normal water level, water depth or channel-bed line. It varies with northing rather than exact chainage along the curved channel. The profile rises sharply about 850–1000 m south of the model's northern edge. Possible causes include structures, the representation of water or bed elevations, or DEM processing. The cause is unverified; this feature must not be interpreted as a measured water-level drop.

For each offset Δh:

`candidate low ground = valid DEM cells where terrain Z < reference Z + Δh`

Only cells connected to river seeds through four-neighbour connectivity are retained. Isolated depressions are not automatically classified as river inundation. Scenario depth is reference elevation plus the offset, minus terrain elevation. The 17 offsets are 0, 0.5, 1.0 … 8.0 m. **The 0 m scenario does not represent observed normal river conditions.**

The method omits conservation of water volume, flow resistance, flood-wave propagation and hydraulic structures. Fine raster resolution does not establish accurate flood prediction. See Sanders, Wing and Bates (2024), [Flooding is Not Like Filling a Bath](https://doi.org/10.1029/2024EF005164), for limitations of static inundation methods.

### Results within the boundary

“Site land covered” excludes the existing mapped water polygon. It is not an estimate of newly occurring damage, asset loss or vegetation mortality. Percentages use the full 68.206 ha site, including mapped water, as the denominator. Areas use 1 m cell-centre membership and are not survey-grade measurements.

| Reference offset Δh | Scenario coverage including mapped water | Site land covered, excluding mapped water | Share of total site |
|---|---:|---:|---:|
| +2 m | 4.63 ha | 2.36 ha | 3.5% |
| +4 m | 6.98 ha | 4.65 ha | 6.8% |
| +6 m | 13.13 ha | 10.81 ha | 15.8% |
| +8 m | 30.22 ha | 27.89 ha | 40.9% |

These offsets **are not rises above today's measured water surface and do not represent 20-, 50- or 100-year floods**. Sudden area changes reflect raster connectivity under this method, not established thresholds in actual flood behaviour.

### Sensitivity to the reference assumption

The reference was also constructed using the 10th and 50th percentiles, with all other method choices unchanged. The resulting ranges across the three reference assumptions are:

| Δh | Range of calculated site land coverage |
|---|---:|
| +2 m | 2.31–2.51 ha |
| +4 m | 4.56–4.92 ha |
| +6 m | 10.60–11.58 ha |
| +8 m | 27.18–29.48 ha |

These are **method-assumption sensitivity ranges**, not confidence intervals. They exclude uncertainty in flow, structures, terrain age and other model inputs. See `Tables/reference_assumption_sensitivity.csv`.

### Interpreting rising and falling stage

The animation steps through increasing offsets and then reverses the sequence. It has no real time axis and does not simulate residual ponding, blocked drainage, groundwater recharge or bank erosion. The same offset produces identical rising and falling footprints because the method is static, **not because real river flooding behaves this way**.

Independent rainfall-driven surface-water flooding and drainage-network surcharge were not simulated. Uncovered land, NoData areas and isolated depressions must not be described as safe. Without building-floor elevations and asset data, the model cannot estimate damaged buildings, economic losses or safe evacuation routes.

### Comparison with the official flood study

The official 2020 report and figure volume are included:

- `02_Flood_Study_Figures_2020.pdf`, Figure 2.2, PDF pages 7–8: 1% AEP flood extent and depth.
- Figure 2.7, PDF page 15: CBD local-flooding mapping; read the legend and treatment of main-channel flooding carefully.
- Figure 2.4, PDF pages 11–12: design water-surface profiles.
- Figures 2.5 and 2.6, PDF pages 13 and 14: bridge-related rise times and rates of rise.
- Figure C2.7, PDF pages 69–70: historical flood-water-surface information.

`Maps/Official_2020_Figure2_2_PDFpage7.png` is a reading copy of the complete original page. It was not digitised into a GIS boundary or used to calibrate the scenarios. [Official figure volume](https://www.qprc.nsw.gov.au/files/assets/public/v/3/building-and-development/planning-docs/flood-risk-management-plans-and-studies/queanbeyan-floodplain-risk-management-study-and-plan-2020-figures.pdf).

A model of velocity, arrival time and recession lag would require the official TUFLOW model/results or consistent-datum channel sections, bathymetry, structures, defences, boundary conditions, flow hydrographs, rainfall, roughness and calibration observations. The [2020 TUFLOW results catalogue](https://flooddata.ses.nsw.gov.au/related-dataset/qfrms-p-tuflow-model-output) was located, but access and third-party licence conditions apply. Its model grids were not obtained; this package does not reproduce the official model.

## 3. What the biological records can establish

### Filtering and coverage

The GBIF occurrence API query and publisher metadata are retained. The initial bounding-box search returned 4,093 records. Removing 658 outside the actual 2 km site buffer left **3,435 records**. The process checked occurrenceID duplicates, future dates and generalised or withheld-coordinate flags; these checks did not remove additional records in this retrieval.

Filters require dates from 2020 through 2026-09-20, coordinates, PRESENT status, no GBIF geospatial-issue flag, and reported coordinate uncertainty between 0 and 100 m. Dates and source-record links are preserved. This uncertainty filter excludes many records with missing or larger uncertainty values, so this is **not a complete inventory of all available biological observations**.

| Measure | Result |
|---|---:|
| Fauna records | 2,951 |
| Flora records | 479 |
| Fungal records | 5 |
| Distinct taxon labels | 371 |
| Labels identified at species rank | 327 |
| Reported coordinates within the site | 212 records / 60 taxon labels |

A record is not necessarily a different individual. Animals can be recorded repeatedly, and different publishers may share observations. Grid summaries do not correct for survey effort, season or detectability. Genus-level and species-level labels may coexist. Do not interpret the 60 site labels as exactly 60 species, or the darkest cells as the places with the highest actual biodiversity.

### River fauna to investigate further

| Taxon | Records within 2 km | Reported coordinates within site | Research use |
|---|---:|---:|---|
| Platypus, *Ornithorhynchus anatinus* | 28 | 3 | River-ecology narrative and bank continuity; not burrow locations |
| Rakali, *Hydromys chrysogaster* | 18 | 7 | Water–land interface and bank structure; not population size |
| Black swan, *Cygnus atratus* | 20 | 17 | Evidence of water-surface use |
| Pacific black duck, *Anas superciliosa* | 28 | 23 | Waterbird observations and shoreline use |
| *Crinia signifera* | 482 | 14 | Frog observations; high counts may reflect repeated monitoring |

The latest retained platypus and rakali records are from August 2026. Check individual GBIF links in `Tables/species_occurrence_records.csv` rather than citing summary counts alone. Membership of the site is based on reported coordinates; uncertainty of up to 100 m may change that classification. The CSV also indicates whether the uncertainty circle intersects the site.

Only five retained plant records have reported coordinates within the site: *Taraxacum officinale*, *Rhaphiolepis bibas*, *Capillipedium spicigerum*, *Brachychiton populneus* and *Trifolium repens*. These cannot represent the actual plant community or replace a tree inventory, canopy assessment or botanical survey. Individual tree species, locations and numbers were not inferred, and occurrence records were not automatically classified as native species.

### Supporting literature

QPRC's [2024 Platypus & Rakali Citizen Science Program Annual Report](https://www.qprc.nsw.gov.au/files/assets/public/v/1/waste-and-environment/environment/2024-platypus-and-rakali-sightings-report.pdf) is included. Its two pages cover reporting across the local government area, with Queanbeyan observations along the Queanbeyan River. A decline in reported sightings must not be converted directly into a population decline or a Site02-specific trend.

In `03_River_Riparian_Platypus_Strategy.pdf` (2012), printed pages 23–27, corresponding to PDF pages 32–36, provide community, platypus and rakali context. Figure 6, PDF page 33, shows then-current regional vegetation and approximate species locations. Historical literature, recommended restoration plants and old threat categories are not a 2026 site survey. [Original strategy](https://www.qprc.nsw.gov.au/files/assets/public/bulding-and-development/planning-docs/plans-of-management/pom_qbyn_river_riparian_corridor_strategy_july_2012.pdf).

## 4. Vegetation communities and river history

The NSW DCCEEW extant and pre-clearing PCT services were downloaded and clipped. The data portal identifies the extant release as C2.0.M2.2 (December 2025). The online service retrieved on 2026-09-20 did not return a release identifier, so its deployment could not independently be matched to that portal version. [Dataset description and licence](https://www.data.nsw.gov.au/data/dataset/nsw-state-vegetation-type-map).

| Category within site | Extant model area | Reconstructed pre-clearing area |
|---|---:|---:|
| PCT 3376: Southern Tableland Grassy Box Woodland | 4.64 ha | 59.83 ha |
| PCT 4085: Southwest Tableland Gorges Riparian Shrubland | 3.63 ha | 8.38 ha |
| Unclassified | 59.93 ha | — |

The extant PCT map is a regional native-vegetation product. “Unclassified” does not mean treeless, unvegetated or impervious. Pre-clearing vegetation is a reconstruction, not a historical survey sample suitable for direct, precise loss-rate calculation. Retain the definitions, classifications and uncertainty of both products. Clipped polygon areas do not establish statutory ecological-community boundaries.

A useful historical sequence is **pre-clearing riparian–woodland relationships → urban development and bank management → present native-vegetation patches and animal observations**. Cross-check the earlier aerial-photo index and bridge/weir history. Distinguish vegetation removal, species change and vegetation that was simply not classified by a map.

## 5. Connecting flooding and ecology

The main scenarios were intersected with classified extant vegetation inside the site. The `Vegetation_Overlap_*` layers show where these mapped features coincide. They do not estimate mortality or permanent habitat loss.

| Δh | Overlap with PCT 3376 | Overlap with PCT 4085 |
|---|---:|---:|
| +2 m | 0.02 ha | 1.72 ha |
| +4 m | 0.31 ha | 2.41 ha |
| +6 m | 0.86 ha | 2.87 ha |
| +8 m | 2.63 ha | 3.35 ha |

These intersections include all overlap between community polygons and scenario coverage, without further removing mapped water. They therefore use a different definition from “site land covered” above. Without inundation duration, velocity, phenology and species-tolerance evidence, ecological damage rates cannot be estimated.

Field investigation can be organised around these questions:

- **Banks and slope toes:** document materials, erosion/deposition, exposed soil, vegetation layers and hard edges; investigate the marked reference-profile change.
- **Tributary confluences and drainage outlets:** verify Buttles Creek and road-drainage connections, distinguishing river backwater from local rainfall ponding.
- **Riparian continuity:** use PCT polygons to select survey reaches, then verify tree species, shrub layers and gaps. Community polygons are not tree crowns.
- **Platypus and rakali:** use consistent public observation points and times; record effort, date, weather and river level so repeat reporting is not mistaken for population change.
- **Frogs and waterbirds:** relate observations to season, local water and bank structure. A single non-detection does not establish absence.

## 6. Licences and quality checks

The retained GBIF records include 2,016 under CC BY-NC, 1,416 under CC BY and 3 under CC0. **The package cannot be assigned a blanket unrestricted commercial licence.** Follow each record's licence and attribute original datasets. The API retrieval is not a formal GBIF download with a download DOI; none has been invented. Publisher metadata and dataset DOIs appear in `Sources_and_Licenses.md`.

FileGDB vectors and GeoTIFFs use EPSG:28355; exported GeoJSON uses WGS84. Raw service JSON, queries, fields, methods and checks are retained in `Metadata`. See `Metadata/GIS_Layers_and_Fields.md`, `project_qa.json` and `english_edition_qa.json`.

The browser security policy previously blocked the local HTML URL. No alternative browser route was used. The English page was checked for resources, links, language and JavaScript syntax, but was not click-tested in the browser. PNG/SVG maps and the GIF are independent of the page.

No source project was modified. Recommendations in reference documents were treated as source material, not as instructions to implement restoration works.


## People, transport and flood exposure

Open the [detailed interactive subpage](Site02_People_Transport_and_Flood.html) from the main atlas. It includes Census 2021, assumed daily activity, mapped transport, concept improvement anchors and potential flood-overlap locations. [Methods and GIS guide](Mobility_and_Flood_Guide.md).