UG21 Terrain Reports: Flow Accumulation Is Not Water Discharge
A flow-accumulation raster can count contributing upstream cells without measuring any water. In a UG21 terrain-mapping inquiry, ask what was accumulated, which weights and routing method were used, and what the output units mean. A drainage-screening map should not acquire a discharge label simply because its largest values follow plausible channels. Specify the intended decision before comparing aircraft or processing proposals.
A column called flow leaves the main question unanswered
Picture a site report with a colored terrain image and a table headed flow. The purchasing manager assumes the largest number describes the amount of water passing a location. The analyst means the number of raster cells contributing through the selected routing model. Both people recognize the same valley in the image, so the misunderstanding can survive an otherwise detailed review. A clear map can conceal an unclear quantity.
This is an editorial scoping example, not a reported UG21 deployment. It illustrates why a deliverable description should include more than a recognizable output name. Ask the proposed analyst to complete this sentence: each value in this file represents a specified quantity, calculated from specified inputs, for a specified use. If that explanation changes between the quotation and the final report, the buyer needs to approve the change.
The software definition does not describe a discharge observation
Esri's Flow Accumulation documentation explains that the tool accumulates upstream cell weights. Without an input weight raster, a weight of one is assigned to each cell, and the output counts contributing cells. That bounded definition supports a useful distinction: the presence of a flow-related tool name does not, by itself, establish a measured volume of water per unit time.
Do not infer a supported UG21 processing integration from that documentation. A supplier might propose another application, a different method or a different product altogether. The contract should identify the actual workflow and its output meaning. Naming a familiar software tool is helpful only when the responsible processor explains how the selected inputs and settings answer the customer's question.
Begin with a screening question
A site team may want to identify terrain locations that deserve closer drainage review. That is a different purchase from a design-ready hydraulic assessment or a record of observed discharge. Write the screening question in ordinary language, such as which candidate locations should receive a specialist's attention. Then identify what the map is explicitly not intended to decide. This keeps the preliminary product useful without allowing its interpretation to expand unnoticed.
Ask who will receive the output. A facilities coordinator may need a marked overview and a list of unresolved questions. A technical analyst may need the underlying raster and processing record. A design professional may require evidence outside the mapping scope. These recipients should not be treated as interchangeable simply because they can all open a PDF. Define the handover around their separate decisions and review responsibilities.
Keep the quantity attached to its legend
The legend is part of the specification, not a cosmetic detail to settle after processing. If the output counts cells, say so. If the analyst proposes weighted accumulation, request the identity, units and rationale of the weights. Ask which interpretation is justified and which interpretation remains unsupported. Avoid an unlabeled progression from low to high when the recipient needs to understand a quantity rather than just a visual ranking.
Our editorial review lesson is to select one output location and ask the analyst to trace its meaning back through the inputs. The buyer is not being asked to reproduce the entire algorithm or approve a hydraulic design. The purpose is to check that the explanation, dataset properties and report wording agree. A discrepancy at one location is a reason for clarification before the label is repeated across a whole site.
| Specification item | Question for the processor | Record to retain |
|---|---|---|
| Output quantity | What does one value represent? | Plain-language definition and units |
| Weight layer | Was a layer supplied or a default used? | Input identity and interpretation |
| Routing method | Which method and version produced this output? | Processing record |
| Decision limit | What must not be concluded from the map? | Approved scope and limitations |

Describe the input terrain, not only the final picture
The receiving team should know which terrain dataset was used and which version belongs to the report. Ask for a concise description of preprocessing and excluded areas, with the original input retained where the agreed scope permits. A buyer should not have to reverse-engineer the source from the appearance of the final map. The important procurement issue is traceability: can the processor identify what changed if the output is regenerated?
Keep acquisition, processing and interpretation evidence in separate parts of the handover. A successfully opened raster is evidence of file usability, not automatic evidence that the source terrain is suitable for every drainage decision. Similarly, a smooth-looking surface does not settle the status of a gap or a changed input. The appropriate specialist should define the evidence required for the intended use before acceptance criteria are finalized.
Make a sample review small and repeatable
Request a bounded sample before commissioning the full reporting package. Agree which file, locations and legend statements the reviewer will inspect. Include an ordinary location as well as one that raises a scope question. Have the recipient record what the values mean in their own words. This can reveal an interpretation mismatch that a general presentation would not expose, without pretending that a small sample validates an entire project.
Record who may approve a limitation and who must escalate it. For example, a presentation editor can correct a mislabeled caption, but should not decide that a screening output is suitable for an engineering calculation. A technical reviewer may accept the method description while still asking for further source evidence. Separating those decisions makes the acceptance conversation more precise than a single unqualified approved stamp.
Do not combine unrelated accuracy claims
A proposal may attach several impressive-looking numbers to a mapping package. Ask what each number evaluates. A classification score, a measurement uncertainty and a processing output value are different categories of evidence. None should be used to fill an empty explanation for another. The UG25 guide to training and validation evidence addresses one common example: a headline score needs the identity and role of the evaluated records.
When a supplier presents measurement evidence as part of the proposed configuration, keep its stated conditions and coverage with the number. The UIV2200 uncertainty comparison guide explains that separate procurement question. These distinctions do not make a terrain screening package unnecessarily complicated. They prevent the report from making several different promises through one ambiguous phrase such as accurate flow mapping.
Put UG21 in a bounded configuration discussion
The approved UG21 product listing identifies the aircraft being discussed. It does not establish that a particular terrain sensor, software license, processing service, drainage analysis or specialist sign-off is included. Request those details as a proposed scope of supply. For unconfirmed specifications and integration requirements, contact us for configuration details. Do not use an illustrative landscape as evidence of aircraft suitability for the site.
Buyers in the United States and the United Kingdom can use the same quantity-definition questions while keeping local operational permissions and professional responsibilities separate. This article does not prescribe flight procedures or establish a jurisdictional approval. Its purpose is to improve the data request so that aircraft suitability, acquisition planning and the later interpretation can each be reviewed by the appropriate person.
Questions before accepting the report
Can a high accumulation value identify a review priority?
It may be relevant to a defined screening workflow, but ask the responsible analyst to explain the interpretation and limitations. Do not let a preliminary priority become an engineering conclusion merely because it has been placed in a final presentation.
Should a buyer demand one particular application?
Specify an application only when it is a genuine receiving-workflow requirement. Otherwise, ask for the quantity definition, method record and inspectable output. Equivalent-looking screenshots are not sufficient evidence that two methods produce an equivalent deliverable.
What should happen when the source or weights change?
Request a revision note identifying the changed input, affected outputs and required review. Preserve the previous scope and state whether the interpretation changed. A replacement file with the same name should not silently replace an accepted meaning.
Take a legend and a decision boundary to the next conversation
Prepare a short brief containing your intended screening question, a draft legend and the decisions that remain outside the mapping scope. Add the expected receiving format and the reviewer responsible for interpretation. Compare the available VTOL and fixed-wing drone options only after these output needs are clear, rather than using an aircraft model name as a substitute for a deliverable definition.
Use the UNITED UAV contact page to discuss UG21 configuration and whether the required acquisition and delivery scope can be supported. Ask for inclusions, exclusions and unresolved evidence in writing. A useful proposal should make it possible to explain what its flow map means without turning a counted-cell output into a water measurement.