How many years of airspace do you really have? The capacity report and the survey behind it
November 18, 2026

Landfill operators sell one product, airspace, and the question every year is how many years of it are left. The answer lives in the annual capacity report, and it is only as good as the survey behind it.
Olinda Alpha's Solid Waste Facility Permit says the operator "shall submit a Capacity Report (hard-copy) no later than February 15 of each year." For a survey, that means flights in December and January, while the previous year's fill is still easy to separate from this year's.
This post walks through what that report asks for, how the volume behind it becomes years of airspace, and how much of the answer rests on survey error. It includes a worked example with made-up numbers, clearly labeled, and a calculation you can rerun with your own area and RMSE.
What the capacity report contains
The Olinda permit lists the minimum content for the preceding year, "with calculations, sources of input data, and assumptions identified." The minimum items are below. Items 2, 3, 5 and 6 depend on a survey.
- Total waste tonnage buried (from scale records, not the survey).
- Volume of airspace filled, in cubic yards.
- Remaining airspace capacity at the end of the reporting year, in cubic yards.
- The airspace utilization factor (AUF) for the reporting year and the long-term AUF.
- A color-coded site plan showing disposal limits, soil stockpiles, and existing and ultimate (permitted) fill contour lines.
- Color-coded cross sections showing existing and permitted fill limits, elevations, stockpiles and property lines.
Permits differ. Borrego's permit is a small-site example. West Miramar's permit on file puts remaining volumetric capacity in a quarterly submittal, with the October to December report due February 1. Read your own permit and your LEA's conditions before writing a scope.
How a surface-to-surface volume is computed
Airspace filled is the volume between two surfaces: last year's year-end ground and this year's. Remaining capacity is the volume between this year's ground and the permitted final grade.
The mechanics are simple.
- Put both surfaces on the same horizontal datum, vertical datum and grid. A geoid model change between epochs can look like fill.
- Subtract: dz = z(new) minus z(old) at each cell.
- Multiply each cell's dz by its area and sum. Keep fill (dz above zero) and cut or settlement (dz below zero) as separate totals.
- Clip to the limit of waste. Stockpiles, haul roads and borrow areas belong in their own polygons because the permit asks for stockpiles separately.
- Check the result with end-area volumes from the permit cross sections. Large disagreement usually means a datum, clipping or interpolation problem, not a measurement problem.
State whether you used a TIN or a grid, and the cell size.
Two cautions. The new-minus-old volume also includes settlement of the waste below, so it is not purely new waste. And remaining capacity measured to a design surface depends on what that surface represents, for example top of final cover or top of waste. Confirm the definition against your permit and closure plan.
A worked example: from AUF to years of airspace (illustrative numbers)
AUF is tons placed divided by airspace consumed. The numbers below are invented for a hypothetical "Example Canyon Landfill." They are not from any real site.
| Item | Illustrative value |
|---|---|
| Waste tonnage buried in the year (scale house) | 1,260,000 tons |
| Airspace consumed (new surface minus old surface, within limit of waste) | 1,050,000 cy |
| Reporting-year AUF | 1,260,000 / 1,050,000 = 1.20 tons/cy |
| Remaining capacity at year end (surface to permitted final grade) | 14,700,000 cy |
| Long-term AUF (cumulative tons over cumulative airspace since the base survey) | 1.15 tons/cy |
Now project the life of the site at the same annual tonnage.
- At the year's AUF of 1.20: 1,260,000 / 1.20 = 1,050,000 cy a year, so 14,700,000 / 1,050,000 = 14.0 years.
- At the long-term AUF of 1.15: 1,260,000 / 1.15 = 1,095,652 cy a year, so 14,700,000 / 1,095,652 = 13.4 years.
The gap is about seven months from one input choice. That is why the permit asks for both AUFs and why the survey behind the year's number matters.
How survey error becomes volume error
Every elevation has an error. The question is how much volume error that creates over a given area. Figure 1 computes it.
Assumptions.
- Each surface has the same vertical RMSE.
- For one point, the uncertainty of a difference of two surfaces is sqrt(a² + b²). With equal RMSE that is 1.414 times the RMSE.
- Volume uncertainty then depends on how the errors are spread over the area A.
Systematic error (a bias that is the same across the whole area, such as a control or geoid offset) does not average out: uncertainty equals sigma times A. Random error that is independent between blocks of side L averages down: uncertainty equals sigma times the square root of A times L. The truth for a real survey sits between those, and the correlation length is an assumption (the 30 m case in the figure is an inferred value, not a measured one).

Some values from the computed table, for a 25-acre cell:
- At 2.5 cm per surface, the bias bound is about 4,700 cy and the 30 m random case is about 440 cy.
- At 5 cm per surface, the bias bound is about 9,400 cy and the 30 m random case is about 880 cy.
- At 7 cm per surface, the bias bound is about 13,100 cy.
Now apply it to the example site. Take a 60-acre active footprint and 5 cm RMSE on each surface.
- Bias bound: 1.414 x 0.05 m x 242,811 m² = 17,170 m³, or about 22,500 cy. That is 2.1% of the 1,050,000 cy consumed.
- That moves the reporting-year AUF between roughly 1.18 and 1.23 tons/cy.
- Random case (30 m blocks): about 1,400 cy, or 0.13%.
- Remaining capacity measured to a design surface carries only one measured surface: 0.05 m x 242,811 m² = 12,141 m³, about 15,900 cy, or 0.1% of 14.7 million cy.
Three things follow.
- Bias is the number to manage. Random noise in a one-meter grid is almost irrelevant at landfill scale. A shared control error or a vertical datum offset is not.
- Common-mode error cancels, up to a point. If both surveys use the same control, datum and sensor setup, much of the bias is shared and drops out of the difference. The sqrt(a² + b²) formula assumes independent errors. If you can show common control, your real uncertainty is lower. If the two years used different vendors or datums, it can be higher.
- Survey error is usually smaller than the other inputs. In the example, the AUF choice moved the life estimate by seven months. A 2% survey bias moves it by less than that. Tonnage and compaction assumptions often matter more, but they are not something a survey can fix.
ASPRS's Positional Accuracy Standards, Edition 2 express vertical accuracy as RMSE, and the standard calls for at least 30 checkpoints. It also says vegetated vertical accuracy should be "evaluated and reported" rather than used as a pass/fail criterion. Ask your surveyor to report RMSEv from independent checkpoints, with the count, and to say where the points were.
Why closure dates move
Closure dates in public records are estimates. They change when a new survey, a new tonnage forecast or a new AUF replaces the old one.
CalRecycle's staff report of February 18, 2026 on the Simi Valley Landfill lists the estimated closure year changing from 2063 to 2056 and the permit incorporating an updated Joint Technical Document. The report does not say why the date moved or which survey it used. SWIS now lists remaining capacity of 72,000,000 cy as of 9/30/2025, with an estimated cease date of November 2056 for the site.
The Toland Road Landfill is a clearer case. Its 2021 public notice proposed moving the closure date from 2027 to April 2033, "based on historical data calculated using the most recent topographic map of the landfill from June 11, 2020." The survey date sits right in the permit.
Public records also lag. Figure 2 plots every active, permitted landfill in the six Southern California counties that lists remaining capacity in SWIS. Of the 29 plotted, 20 show a capacity date more than three years before October 17, 2026.

Two cautions on reading it.
- SWIS gives no method or source for these figures. A stale date may mean the number is old, or that the operator reported it to the LEA and CalRecycle was never updated. The chart cannot tell which.
- The two public numbers for the same site can disagree. Olinda's permit cites 20,684,131 cy from the Annual Capacity Report for 2020. SWIS shows 17,500,000 cy as of 10/1/2020. The public documents reviewed for this post do not explain the gap. Different definitions of capacity are one possible reason.
Chiquita Canyon is left off the chart on purpose. SWIS lists it as active, but the operator's site says that as of January 1, 2025 the landfill is closed for the acceptance of waste. Its SWIS remaining-capacity date is also from 2021, so the figure no longer describes an operating site.
Flight logistics on an active landfill
The following is practical reasoning about planning, not a regulatory checklist. Confirm site rules with the operator.
- Schedule around the working face. Compactors, trucks and spotters are on it all day. Plan flights for off-hours, or fly the active face in a short window while equipment is held back. Part 107 restricts flight over people who are not directly participating, unless they are under cover (14 CFR 107.39).
- Plan altitude to the terrain. The Part 107 limit is 400 feet above ground level, with an exception near structures (14 CFR 107.51). A 300-foot-tall fill makes "ground level" a moving target, so use terrain-following paths rather than a flat altitude.
- Control goes where it will survive. Targets cannot sit on a face that gets rebuilt daily. Set them on perimeter roads, pads and stable cover. Set independent checkpoints in the same places, plus on soil-covered slopes inside the limit of waste.
- Fly after the last fill of the period, not the first. Waste placed after the flight lands in next year's number.
- Allow weather days. Winter storms push flights, so build in slack before the deadline.
LiDAR or photogrammetry on vegetated closed slopes
On a bare active face, photogrammetry with good control can give usable volumes. On closed, vegetated slopes the two methods diverge.
Photogrammetry reconstructs what the camera sees, which is the top of grass, brush or trees. A bare-earth surface has to be inferred. LiDAR records some returns that reach the ground between plants. That does not make it perfect. Tamimi and Toth tested a DJI Zenmuse L2 and found an RMSE of 0.07 m on an open road and 0.21 m under dense canopy, a threefold change. The Chen paper reports 2 to 3 cm over bare ground. Vegetation on closed slopes can push a LiDAR surface toward the canopy figure.
For closed sites, 27 CCR 21090(e) asks for a five-yearly iso-settlement map. A bias of 10 or 20 cm from vegetation can mask much of the settlement you are trying to measure. Check points on the vegetated slopes themselves, and report that accuracy separately from the open-ground figure.
Who has to sign
A topographic surface and volumes delivered for a permit are land surveying under California law. Business and Professions Code 8726 includes anyone who "determines the configuration or contour of the earth's surface" using mathematics or photogrammetry. The Board's Bulletin 50 says flight planning, piloting and capturing photos are not themselves regulated, but creating a topographic survey or an orthorectified map for a client or public agency is. That work needs a licensee in responsible charge: a PLS or, in California, a civil engineer legally authorized to practice land surveying.
Twinn Digital captures and processes data. Flights are planned with software agents and flown by a human FAA Part 107 remote pilot, on equipment rented for the job. Twinn does not stamp. A licensed partner reviews and stamps what goes into a permit submittal. The documents reviewed for this post do not say who signs the annual capacity report at any given site, so ask the engineer of record early.
Federal-land exceptions: Miramar and Camp Pendleton
Federal drone rules turn on federal funding or federal operation, not on the landfill itself. Two sites in the region need a closer look.
- West Miramar. The City of San Diego operates it, but the permit lists the owner as the United States, Department of the Navy, MCAS Miramar (SWFP). If the work is paid from city funds, the federal funding ban does not follow from ownership alone. That is an inference, not legal advice. Base access and airspace approvals are another matter and need a base contact.
- San Onofre Landfill, Camp Pendleton. SWIS lists the US Marine Corps as operator. When a federal agency operates or contracts for drone services, the American Security Drone Act (Pub. L. 118-31, secs. 1821-1833) bars use of drones from covered foreign entities starting December 22, 2025, and sunsets December 22, 2028. Before bidding federally funded or federally operated work, check whether a manufacturer such as DJI is listed as a covered entity on SAM.gov rather than assuming either way. NDAA-compliant sensors are available on request for agency-facing work.
The FCC's December 22, 2025 Covered List action is a different mechanism. It blocks new equipment authorizations, not the use of drones already owned.
What the terrain looks like
Figure 3 shows Olinda Alpha from public USGS 3DEP lidar. It is a good reminder of what public data does and does not give you.

The bench geometry is clear. But the tile is eight years old, and an active site changes every month. The public surface can serve as a historical baseline. It cannot stand in for a current survey or a stamped deliverable.
One practical next step
Before the December flight window, collect four things from your engineer of record:
- Last year's capacity report and the surface it used.
- The datum, grid and stationing the permit cross sections follow.
- The control monuments used last year.
- The definition of the final-grade surface used for remaining capacity.
Then settle one question in writing: will the new survey be tied to the same control, and who reports RMSEv from independent checkpoints? Those answers decide how much of your uncertainty in Figure 1 is bias and how much is noise.
To plan a December to January capture and processing job with a licensed partner stamping the deliverable, see Twinn Digital for landfills.