Read Pittsburgh River Gauge Products Without Mixing Stage, Flow and Time

A Pittsburgh river chart can put a water-level reading, a flow estimate, a forecast and a timestamp close together. They answer different questions. Before repeating a number in a neighborhood discussion, research note or outing-planning conversation, identify the location, quantity, reference and time attached to it. A number detached from those labels can acquire a meaning that the original product never supplied.

This is a guide to reading the structure of river-gauge information. It does not report current river conditions, assess a route or determine whether a water activity is suitable. The National Weather Service Pittsburgh hydrology page provides a local starting point and links to site hydrographs through the National Water Prediction Service. Use the current official products and instructions for current decisions. The numerical cases below are invented teaching examples, not measurements from Pittsburgh rivers.

Start with the location and the product

Open the official source rather than relying on a cropped chart shared without its heading. Write down the monitoring location name and identifier exactly as displayed. A river name alone may cover many locations. A reading associated with one location should not silently become a statement about every riverfront, bridge, tributary or trail bearing the same river name. The location is part of the observation, not an optional caption.

Then identify what the product presents. A regional table can be a discovery tool for several locations. A single-site hydrograph can show values over time. A forecast product can extend a series into a future period. Those differences affect what comparisons are available. A regional summary is not automatically a substitute for the detailed site information, and a chart showing a time series is not automatically a complete record of every condition someone might care about.

The Pittsburgh hydrology page describes hydrographs as showing stages and flows over time and directs readers to site names for those charts. This guide uses that page for discovery context only. It does not reproduce its changing numerical rows as a current conditions report. If you take notes from a live product, retain the source link and the time you accessed it so another reader can return to the appropriate official page.

A useful note has several separate fields: source, site identifier, parameter, unit, observation or valid time, product issue time where supplied, and status or qualification. Keeping these fields separate makes an explanation easier to check. A sentence such as the river is at five leaves almost all of them unresolved. A careful reader should be able to identify what five measures before deciding what the sentence supports.

Stage describes a water-surface reference, not a universal depth

USGS explains gage height as the water surface's height above a reference datum selected for that gauge. NOAA's hydrograph terminology similarly defines stage relative to a datum. These definitions explain why a stage reading should not be casually renamed water depth. The reference zero is part of the measurement definition. It is not automatically the riverbed at the place a person wants to stand, walk or paddle.

A datum is easiest to understand as the origin of a vertical ruler. Moving the ruler's zero changes the number assigned to the same height. That simple relationship is useful for interpreting labels, but it does not authorize a reader to combine actual gauges using guessed reference elevations. A valid comparison needs documented, compatible reference information for the specific sites and records involved.

Stage is also different from water-surface elevation expressed in a common vertical reference. Two products can use feet while measuring relative to different zeros. The shared unit does not make the numbers directly comparable. Read the parameter and datum notes before subtracting values. When the reference information is absent from an excerpt, say that a cross-site height comparison remains unresolved rather than constructing a conversion from intuition.

Do not infer a channel's depth from a stage alone. That would require information about the relevant bed elevation and location, along with the relationship between the available reading and that location. This guide has not inspected those things. A label-reading exercise can identify why the inference is unsupported without making an alternative claim about actual depth or navigability.

Case A: the same height can have different stage numbers

Imagine two deliberately fictional vertical rulers used for the same teaching diagram. Ruler A has its zero at 100 invented height units in a common reference, and the water surface is five units above that zero. Ruler B has its zero at 98 units and the same surface is seven units above it. Both calculations give a common-reference height of 105: 100 plus five, and 98 plus seven.

The stage numbers differ by two, while the calculated common-reference heights in this authored diagram differ by zero. A larger stage number therefore does not, by itself, establish a higher common-reference water surface when the zeros differ. The arithmetic illustrates the role of the reference. It is not a conversion between real Pittsburgh gauges and does not assert that any particular official datum has either of these values.

A practical comparison note should state whether the two values use the same site, parameter, datum and unit. If they do not, identify the missing relationship before calculating. If a source documents a datum change, retain the effective period and the source's conversion instructions. Do not splice two historical series together merely because their plotted lines appear to connect smoothly in a screenshot.

When explaining the example to someone else, name both additions. Saying that the gauges disagree would be misleading because the authored rulers describe the same height using different origins. Saying that real gauges always can be reconciled by adding a small offset would be equally unsupported. The conclusion belongs to this declared diagram, where the common reference and both zeros were supplied explicitly.

Flow is a volume rate and needs its own label

NOAA defines flow as volume passing a point per unit time. USGS's streamgaging material distinguishes water-level observations from discharge and describes a site-specific stage-discharge relationship. The distinction matters because a vertical level and a volume rate are different quantities. A stage change cannot be converted into a discharge change by treating feet as if they were cubic feet per second.

A reader may encounter discharge reported in cubic feet per second, often abbreviated cfs, or in thousands of cubic feet per second, shown as kcfs in NOAA's terminology. The unit tells you that time is in the denominator and volume is in the numerator. It does not describe an individual water particle's speed. Keep the complete unit in notes and comparisons so a shortened label does not turn a volume rate into a length or velocity.

A site's relationship between stage and discharge should come from the responsible source for that site. This guide does not supply a rating curve, estimate flow from a real water level or extrapolate beyond a documented relationship. The useful reading habit is to find the parameter actually displayed instead of substituting a quantity that feels more intuitive. A number can be accurate for its stated parameter and still fail to answer a different question.

Case B: a stage percentage is not a flow percentage

In the second fictional case, one authored stage changes from four to six units using the same declared ruler. The numerical difference is two units. The second value is one and a half times the first, so the increase relative to the first number is one half, or 50 percent. Those calculations describe the supplied stage numbers under their fixed reference. No discharge values have been supplied.

Calling that a 50 percent increase in river flow would change the quantity without evidence. It would also hide the role of the reference zero in a stage percentage. If the same diagram were relabeled with a different zero, the stage numbers and their percentage change could differ even though the underlying vertical separation remained two units. A percentage needs a meaningful base as well as correct arithmetic.

For a communication note, the two-unit difference is the most direct statement supported by this example. If someone asks about flow, the appropriate next question is whether the relevant official discharge series or documented site relationship is available. Do not insert a guessed multiplier to complete a table. The absence of a supplied flow value is a reason to keep that comparison open.

This distinction also helps when an assistant drafts a summary of a chart. Supply the full axis label and ask it to repeat the quantity and unit in its explanation. An otherwise fluent summary can become misleading if it replaces stage with flow. A bounded review can catch that substitution without claiming that an AI system has been evaluated or that any real summary has already made the mistake.

Case C: a unit conversion leaves the underlying flow unchanged

The third authored example is a flow of two kcfs. Using the stated definition that one kcfs equals one thousand cubic feet per second, the same quantity is 2,000 cfs. The numerical labels differ by a factor of one thousand, while the represented flow is unchanged. This is a unit conversion in a fictional example, not evidence that a real river's flow increased between observations.

If a chart switches its unit label, compare the quantities after applying the correct scale. A reader who copies only the numbers could describe two becoming 2,000 as an enormous change. Retaining the units makes the explanation straightforward. The conversion factor belongs to the units; it does not arise from a change in rainfall, reservoir operation or any other physical cause.

Do not abbreviate either label to feet. Cubic feet per second describes volume per time, and thousands of cubic feet per second retains that same dimension. Neither expression supplies a velocity at a particular point. Answering a speed question would require different information. A clear summary can state the conversion and stop at the boundary of what the supplied quantity means.

For a reusable comparison sheet, give the original value and unit their own columns and put any converted value in additional columns. Record the conversion factor rather than overwriting the source. That makes a scale change visible and lets another reader reproduce the arithmetic. It also avoids treating a converted presentation as an additional independent observation.

Observation time, forecast time and page access time differ

A river page may contain several kinds of time. An observation time identifies when a reported value applies. A forecast valid time identifies the future time for which a forecast value is intended. A product issue time identifies a version of a forecast or message where one is supplied. A page access time records when you retrieved the page. Reading one timestamp does not resolve the others.

Write the time basis exactly as displayed, including an offset or named time zone when it is present. Do not add a Pittsburgh local-time label to a value marked UTC without doing a documented conversion. If a cropped image omits the time basis, the missing context is part of the uncertainty. Guessing it can move a value into the wrong hour or even the wrong calendar date.

Separate a chart's observed and forecast portions using the product's own labels and legend. A continuous-looking line does not make future values observations. A forecast can be useful information while still having a different evidentiary status from a reported observation. Preserve that distinction in the verbs you choose: reported and forecast communicate different relationships to time.

Case D: a displayed value can precede the time you read it

In this fictional case, a value is labeled with an observation time of 08:00 UTC on June 12, 2030. The reader records a page access time of 09:15 UTC on the same date. The gap between those two supplied timestamps is 75 minutes. This calculation does not measure an actual reporting delay, network latency or gauge malfunction. It compares two invented time labels.

The same case includes an invented forecast issued at 09:00 UTC for a valid time of 11:00 UTC. Its valid time is two hours after its issue time. At the page access time of 09:15, the forecast's valid time is still one hour and 45 minutes away. Those relationships explain why the forecast value should not be described as an observation at the time the page was read.

A concise note could retain all four labels: observation at 08:00, access at 09:15, forecast issued at 09:00 and forecast valid at 11:00, all UTC. The words identify what each clock value does. Removing them and listing only a latest number would conceal whether latest referred to an observation, an issued forecast or a retrieved page.

If you compare two forecast versions, hold the site, parameter and valid time constant before describing a revision. A forecast for tomorrow morning and another for tomorrow evening are different targets even if one was issued later. The example teaches timestamp bookkeeping; it does not establish the accuracy of a real forecast or the cause of a change in a real river.

A few points do not automatically make a daily statistic

USGS distinguishes daily data and statistics from individual time-series values. Its statistics documentation explains that daily data generally summarize continuously sampled data for a day and notes that different daily statistics can be published. For a reader, the key question is which statistic the source actually labels. Mean, maximum, minimum and a value at a specified time are different summaries.

The word daily can also describe a display interval or update routine rather than the exact calculation you have in mind. Read the product definition. Do not derive an official daily mean by averaging whatever points happen to be visible in a screenshot. The selected points may not cover the day, may be unequally spaced or may omit values the official calculation uses.

Case E: the mean of four supplied points has a narrow meaning

The fifth fictional case supplies four point values, ten, twenty, thirty and forty cfs, at 00:00, 01:00, 03:00 and 06:00 UTC. Their arithmetic sum is 100 and their equally weighted mean is 25 cfs. There are four points, and the gaps between their timestamps are one, two and three hours. Those results can be checked directly from the supplied teaching rows.

The value 25 is not an official daily mean. The case does not supply a full day of continuous observations or a declared rule for representing the intervals between the points. Treating every point equally is a chosen arithmetic operation on the list. It is not evidence that the river spent equal amounts of time at those four values or that this list captures all relevant variation.

An explanation should therefore say mean of the four supplied point values. That wording identifies the population actually summarized. It leaves the daily statistic unresolved instead of allowing a familiar calculation to stand in for a different product. If a daily mean is needed, look for the responsible source's published daily series and its definition rather than promoting a partial calculation into an official result.

This is also a reason to retain missing and omitted periods in a research note. An unavailable point is not zero flow, and a chart that does not display a point does not necessarily prove the source lacks it. Keep the distinction between the displayed selection and the underlying record. The practical aim is an explanation whose denominator and time coverage a reader can inspect.

Case F: a revised record is not a later physical observation

The final authored case has two versions of the same fictional record at the same site and observation time. Version one gives a value of twenty cfs with an invented provisional label. Version two gives twenty-two cfs with an invented reviewed label. The numerical revision is two cfs, or ten percent relative to twenty. The supplied observation time has not changed.

Calling this a two-cfs rise over time would confuse record revision with a new observation. The exercise changes a reported value for the same target, not the target time. It does not establish why any real data might be revised, and its labels are an invented codebook rather than a claim that every official interface uses those exact words.

USGS documentation warns that provisional data can change and describes examples of adjustments to provisional records. Preserve the status and version context when quoting such a value. A provisional label does not tell you the exact future correction, while a revised value does not by itself reveal the physical cause of river behavior. The two questions require different evidence.

Frequently asked questions

Does a stage of five feet mean the river is five feet deep?

No such depth conclusion follows from that number alone. Stage is referenced to a datum. The bed and location information needed for a depth interpretation are separate. This guide does not assess actual depth, navigation or conditions at any Pittsburgh site.

Can I compare stage numbers from two different gauges directly?

First establish compatible sites, quantities, units and reference information. Shared units do not establish shared zeros. The fictional two-ruler example illustrates the problem but supplies no real gauge conversion or cross-site elevation finding.

Is kcfs a different kind of measurement from cfs?

Both express a volume rate. In NOAA's terminology, one kcfs is one thousand cfs. Two kcfs and 2,000 cfs therefore represent the same flow in the authored example. Neither unit supplies a local water velocity or an activity recommendation.

Is the latest point on a chart necessarily an observation?

Read the legend, timestamps and product labels. A chart can contain observed and forecast portions. Preserve the observation or valid time and the forecast issue time where supplied instead of assuming that the rightmost point describes a measured current condition.

Can I average a few chart points to get the daily mean?

You can calculate the mean of those supplied points under an explicit rule, but that is not automatically the official daily statistic. Establish the coverage and calculation definition, or use the responsible source's published daily product for the question it actually answers.

What do the downloadable cases demonstrate?

They check the arithmetic, units, timestamp relationships and literal version distinctions in six original fictional examples. They do not inspect a real gauge record, evaluate a forecast, estimate river conditions or assess a route, property, water activity or individual safety.