Visualising prior outputs

After a run, emissions layers are written to NetCDF in data/outputs/ (default filename prior-emissions.nc). Flux variables such as ch4_total and ch4_sector_* use the unit kg/m2/s.

Those values are often very small (commonly 1e-11 to 1e-8). That is expected for a per-second, per-square-metre flux, but it makes maps hard to read on a linear scale. Two practical options are a power-of-ten display scale (see below) or converting to another unit such as g/m²/day.

Main variables

Variable	Description
ch4_total	Total methane flux across all sectors
ch4_sector_*	Per-sector flux (e.g. ch4_sector_coal, ch4_sector_transport)
land_mask	Land/sea mask

Display unit conversions

These conversions are for inspection and mapping only. The NetCDF file keeps kg/m2/s.

Let flux be a value in kg/m2/s.

Display unit	Formula
g/m²/day	$\text{flux} \times 86400 \times 1000$
ng/m²/s	$\text{flux} \times 1\text{e9}$

Example: 2.2e-8 kg/m2/s ≈ 1.9 g/m²/day.

Power-of-ten display scale

If you mainly want a readable colour ramp in QGIS without changing the physical unit, multiply the raster by 10^n and note the exponent on the legend. The underlying data stay in kg/m2/s.

Raw max (approx.)	Multiply raster by	Legend label example
1e-8	1e8	×10⁻⁸ kg/m²/s
1e-11	1e11	×10⁻¹¹ kg/m²/s

Pick n so that typical non-zero values land roughly in 0.1–100 on the colour bar. This is a display trick only — use g/m²/day or the raw unit when comparing numbers across runs.

Per-cell totals

To estimate emissions from one grid cell over a day:

$ \text{kg}/\text{day} \text{per} \text{cell} = \text{flux\_kg\_m2\_s} \times \text{cell\_area\_m2} \times 86400 $

Open Methane domains use a rectilinear grid with constant cell size. Cell area is:

$ \text{cell\_area\_m2} = \text{DX} \times \text{DY} $

where DX and DY (metres) are stored as attributes on the output file. For the default aust10km domain, DX = DY = 10000, so each cell is 1 × 10⁸ m².

Python check

import xarray as xr

ds = xr.open_dataset("data/outputs/prior-emissions.nc")
flux = ds["ch4_total"]  # kg/m2/s

g_per_m2_day = flux * 86400 * 1000
cell_area_m2 = float(ds.DX) * float(ds.DY)
kg_per_day = flux * cell_area_m2 * 86400

print("max flux (kg/m2/s):", float(flux.max()))
print("max flux (g/m2/day):", float(g_per_m2_day.max()))

QGIS workflow

Tested with QGIS 3.x against a local aust10km run (prior-emissions.nc, single time step, Lambert Conformal projection).

Load a flux layer

1. Layer → Add Layer → Add Raster Layer
2. Select data/outputs/prior-emissions.nc
3. QGIS lists one subdataset per variable. Pick the one you need, e.g. ...:ch4_total (the exact prefix depends on GDAL/QGIS version)
4. Confirm the layer CRS matches the domain Lambert Conformal projection — WGS84 basemaps should still align with on-the-fly reprojection enabled

For a single-day run the file has one time step; QGIS usually loads it as a single band without extra steps. Multi-day outputs may need the Temporal Controller to pick one date.

Symbology

Raw kg/m2/s on a linear scale hides most structure. Either:

• enable Logarithmic min/max under Layer Properties → Symbology, or
• use a power-of-ten scale (below) so the colour bar shows ordinary-sized numbers

Steps:

1. Layer Properties → Symbology
2. Render type: Singleband pseudocolor
3. Min/max: Min/max or Percentile
4. For log view: enable Logarithmic (wording varies slightly by QGIS version)

Use a simple palette for land_mask.

Power-of-ten scale in QGIS

Raster → Raster Calculator (example for typical ch4_total magnitudes):

"ch4_total@1" * 100000000

Label the legend ×10⁻⁸ kg/m²/s. Adjust the multiplier and exponent label to match your layer (e.g. * 1000000000000$ \text{with} \text{legend} ** \times 10⁻¹¹ \text{kg}/\text{m}²/\text{s}** \text{when} \text{values} \text{are} \text{closer} \text{to} $1e-11).

Convert to g/m²/day (alternative)

Raster → Raster Calculator:

"ch4_total@1" * 86400 * 1000

Label the legend as g/m²/day. Adjust @1 to match the band name shown in the layer list.

Comparing more than one layer

QGIS applies symbology independently to each layer. If you overlay ch4_sector_coal and ch4_sector_transport, each layer is stretched to its own min/max. A saturated colour on transport does not mean it has the same flux as coal — coal values are often orders of magnitude larger.

To compare sectors fairly:

• use the same display unit (e.g. g/m²/day) on each layer, and
• set matching min/max or percentile ranges manually, or
• compare layers one at a time, or
• subtract layers (Raster Calculator) when looking at before/after or difference maps.

Export a GeoTIFF (optional)

To share a map or reload it without selecting NetCDF subdatasets:

1. Load and style the layer as above
2. Project → Export → Export Map to Raster for a map image, or
3. Right-click layer → Export → Save As… to write a GeoTIFF (choose target CRS, e.g. EPSG:4326, if you need WGS84)

There is no built-in prior step that writes GeoTIFFs automatically; export from QGIS or a separate script if you need them.

Common pitfalls

• Linear colour scale on raw kg/m2/s — use log symbology or a power-of-ten multiplier first.
• Overlaying sectors without shared scaling — colours are not comparable (see above).
• Ocean cells — mask with land_mask when inspecting land-based sectors.
• Inventory totals — maps show spatial distribution; integrating flux over area and time is a separate calculation.

See also

• Outputs in the README
• Data sources