AERSURFACE's choices: seasons, moisture, sectors & why they matter

The AERSURFACE page covered the machinery — land cover in, three surface characteristics out. This page covers the judgment calls the tool asks of you, what each one physically changes, and how those changes propagate into modeled concentrations. None of these are obscure knobs: two modelers processing the same station with different season and moisture choices will produce different design values, and a reviewer will ask why.

Re-mapping months to seasons

AERSURFACE's five seasons are land-surface states, and the default month assignments (see the AERSURFACE page's table) describe a generic mid-latitude site. The tool expects you to re-map them to your climate — including the biggest single call:

Does winter have continuous snow cover? The two winter categories carry very different surface values — snow is bright (high albedo) and smooth (low roughness). Assigning December–February to "winter with continuous snow" in a place that actually has bare, dark winter fields misrepresents both the energy balance and the mechanical turbulence for a quarter of the year. Example: a northern-plains site might justify snow cover for December through March; a Gulf Coast site never uses the snow category at all, and its "winter" months belong in "late autumn after frost and harvest, or winter with no snow."

Why it moves results: albedo feeds the daytime energy balance (higher albedo → less absorbed sunlight → weaker sensible heat flux → weaker convective mixing), and roughness sets the friction velocity (smoother surface → lower u* → weaker mechanical mixing in stable hours). Both act on every hour of the affected months.

The surface-moisture condition

Bowen ratio values depend on whether the period was wet, dry, or average — and the User's Guide gives an explicit, checkable recipe:

The surface moisture condition can be determined by comparing precipitation for the period of data to be processed to the 30-year climatological record. It is recommended the user specify "wet" conditions if precipitation is in the upper 30th-percentile, "dry" conditions if precipitation is in the lower 30th-percentile, and "average" conditions if precipitation is in the middle 40th-percentile.

Source · AERSURFACE User's Guide (EPA-454/B-26-007) §2.3.3.

The same section adds two subtleties: the condition applies to the entire data period — "if the surface moisture condition varies significantly across the data period, then AERSURFACE may need to be applied multiple times" (one run per year is the clean pattern for a five-year record with a drought year in it) — and "what is normal varies for different regions," so a categorization that departs from the percentile recipe should be justified.

Why it moves results: the Bowen ratio splits the surface's energy between heating air and evaporating water (the daytime heat flux is H = 0.9·R_n/(1 + 1/B₀) — see the boundary-layer lesson). A "dry" year pushes the Bowen ratio up, the heat flux up, and daytime convection up; a "wet" year does the opposite. For arid-region land covers (Barren, Shrubland, Planted/Cultivated classes), the arid/non-arid setting shifts all three characteristics — "the albedo and Bowen ratio will be higher and the surface roughness lower for arid regions" (§2.3.3).

Sectors and the airport question

Roughness is computed per wind-direction sector, and each sector can be flagged airport or non-airport for the ambiguous "Developed" land-cover categories — judged "by the predominant land use within a kilometer radius of the meteorological tower, giving more weight to land use nearest the tower" (§2.3.2; see the AERSURFACE page). The flag matters because the same National Land Cover Database (NLCD) pixel class can mean runways (smooth) or buildings (rough): at a typical airport station, the sectors facing the runways earn the low-roughness treatment while a sector facing a terminal complex may not.

Why it moves results: each hour's u* is computed with the roughness of the sector the wind came from. Misflagging the sector that dominates your worst-case wind direction changes the mechanical turbulence on exactly the hours that control the design value.

What you should not change casually

The default 1-km roughness radius (the ZORAD method) is a considered default — the User's Guide points to the AERMOD Implementation Guide for "conditions for possible exceptions to this default distance" (§2.4.1), and the alternative ZOEFF method "is considered research grade" (§1.3). Radius changes and ZOEFF belong in the with-the-reviewer category, not the routine-setup category.

A worked contrast

Two runs of the same rural Kansas station, identical in everything except AERSURFACE choices. Run A: default months, "average" moisture. Run B: December–February assigned to continuous snow (the site rarely holds snow), and "wet" chosen because the analyst eyeballed a rainy spring. Run B's winters have brighter, smoother ground (less daytime mixing, less stable-hour turbulence) and its Bowen ratio is lower year-round (weaker convection). Neither run is wrong by inspection — but only Run A's choices survive the §2.3.3 percentile check and a snow-climatology look. The reviewable path is the recipe: climate normals for the snow question, the precipitation percentile for moisture, and documentation for both.

In PlumeSmart

Surface characteristics are generated with the station's climate context visible — seasonal assignments, moisture condition, and sector layout are recorded with the met dataset, so the choices behind every z₀, Bowen ratio, and albedo are part of the reviewable record.