Sea-level rise is measured by combining several kinds of evidence rather than relying on one instrument. This guide explains what tide gauges, satellite altimetry, GNSS land-motion stations, coastal benchmarks, and climate-model projections actually measure, how to compare their records, and how to estimate whether a local coastline is rising faster or slower than the global average.
Overview
“Sea level” can describe different quantities. A tide gauge usually records the height of the water relative to a nearby point on land. A satellite altimeter estimates the height of the ocean surface relative to a global reference surface. A GNSS station measures how the land itself moves. A climate model projects future changes under specified assumptions about emissions, ice sheets, ocean warming, and other processes.
These measurements answer related but different questions. A global mean sea-level trend is useful for tracking broad changes in the ocean system. A local relative sea-level trend is more relevant to a harbor, wetland, road, or neighborhood because it includes both ocean change and vertical land motion. Tides, storms, waves, groundwater extraction, sediment compaction, and seasonal circulation can add short-term variation to either record.
For a clear sea-level rise trends analysis, first identify the quantity being reported. Then check the time period, vertical reference, geographic coverage, uncertainty, and whether the result is an observation or a projection. A number without those details is difficult to interpret responsibly.
| Method | Typical time coverage | Geographic scale | Main uncertainty or limitation | Best use |
|---|---|---|---|---|
| Tide gauges | Often decades or longer, depending on the station | Point or harbor location | Measures water relative to land; local land motion and gaps matter | Long-term local relative sea-level trends |
| Satellite altimetry | Modern satellite-era records, with missions and processing versions varying | Broad ocean coverage | Coastal measurements are more difficult; orbit, instrument, and processing corrections matter | Regional and global ocean-surface change |
| GNSS land-motion measurements | Depends on station history and data continuity | Point location near or at a land station | May not represent motion across the entire nearby coastline | Separating ocean change from land uplift or subsidence |
| Coastal benchmarks | Repeated surveys or construction records; coverage can be irregular | Specific structures, survey points, or flood-control assets | Reference points can be disturbed, moved, or tied to different datums | Connecting water levels to elevations used in maps and engineering |
| Climate-model projections | Future periods selected by the analyst | Local, regional, or global depending on model and downscaling | Scenario, model, ice-sheet, and local-process uncertainty | Planning for possible future conditions |
How to estimate
A basic trend estimate describes the average change in a measured quantity over a chosen interval. If a record has a fitted slope of b units per year, the estimated change over an interval of n years is:
Estimated change = b × n
This calculation is simple, but selecting and interpreting b requires care. A trend is not the same as the water level on a particular day, and extending a historical trend far into the future assumes that the processes controlling it remain comparable. That assumption is often unsuitable for long-range planning.
- Define the question. Decide whether you need a local relative sea-level trend, an ocean-surface trend, or a future projection. For flood planning, local relative sea level is usually the more relevant starting point.
- Find a record near the location. Search a recognized tide-gauge, ocean-altimetry, GNSS, or coastal-elevation data service. Record the station name, coordinates, operating dates, data version, and units.
- Inspect the time series. Look for missing intervals, abrupt shifts, changes in instrumentation, and unusually short records. Seasonal cycles and individual storms should not be mistaken for a long-term trend.
- Check the reference datum. Determine whether elevations are relative to a local benchmark, a tidal datum, a geodetic reference frame, or another vertical reference. Two records can use the same units while measuring from different zero points.
- Account for vertical land motion. Pair the tide-gauge record with GNSS or other geodetic information where possible. If land is subsiding, the water level relative to the land rises faster than the ocean surface alone. If land is uplifting, the local relative trend can be reduced.
- Compare like with like. Match the time period, units, datum, seasonal treatment, and geographic scale before comparing two trends. Do not compare a short coastal record directly with a global mean without explaining the difference.
- Report a range and context. Include the trend units, period, uncertainty or confidence information supplied with the dataset, and the main limitations. Use “estimated,” “approximately,” or similar wording when the evidence does not support false precision.
For background on reference periods, resolution, and uncertainty, see How to Compare Climate Datasets. Tidal cycles can also complicate short observations; Tides Explained provides useful context before interpreting daily or monthly water levels.
Inputs and assumptions
A repeatable estimate should preserve the inputs that produced it. At minimum, record the following:
- Location: station identifier, coordinates, and distance from the site being studied.
- Measurement type: tide-gauge water level, satellite-derived ocean height, GNSS vertical motion, or model projection.
- Time window: start and end dates, including whether the record is continuous.
- Units: millimeters or centimeters per year for trends, and the conversion used if changing units.
- Vertical reference: the datum or reference frame and any transformation applied.
- Trend method: for example, a linear fit, a published trend product, or a model-derived projection.
- Uncertainty: the stated uncertainty around the trend or projection, plus known limitations.
- Scenario and baseline: for projections, specify the scenario, baseline period, target year, and whether the value is global, regional, or local.
Local relative sea-level change can be described conceptually as:
Relative sea-level trend ≈ ocean-surface trend − land-uplift trend
In this expression, land uplift is positive upward. Subsidence is therefore a negative land-motion value and increases the relative sea-level trend. This is a useful framework, not a guarantee that two independently processed datasets can be combined without checking their reference systems and time windows.
Climate-model projections should not be treated as a continuation of a tide-gauge line. A projection represents a conditional future estimate based on assumptions and model results. When comparing observations with projections, align their baseline period and geographic definition, and distinguish a central estimate from a range.
Worked examples
Example 1: Converting a local trend
Suppose a tide-gauge record reports an illustrative local trend of 4 millimeters per year over a defined observation period. A simple interval estimate for 25 years is:
4 mm/year × 25 years = 100 mm
That equals 10 centimeters of cumulative change under the assumption that the fitted trend remains constant. It does not predict the water level on a specific date, include storm surge, or establish that the next 25 years will follow the same rate. A responsible report would identify the station, period, datum, trend uncertainty, and whether vertical land motion was included.
Example 2: Separating land motion
Imagine a coastal tide gauge with an illustrative relative trend of 5 mm per year. Nearby GNSS observations indicate land subsidence of 2 mm per year. Using the sign convention above, the approximate ocean-surface contribution would be:
5 mm/year relative trend − 2 mm/year subsidence contribution = about 3 mm/year ocean-surface trend
This is a simplified decomposition. The stations may not sample the same location, and local groundwater withdrawal, sediment compaction, tectonic movement, or processing choices may affect the result. The example shows why a tide gauge alone cannot always explain the physical cause of a local trend.
Example 3: Comparing a projection with an observation
A planner wants to compare a historical tide-gauge trend with a future local projection. The correct workflow is not to place the two numbers side by side immediately. First, confirm that both refer to the same coastline or a clearly defined nearby area. Next, record each baseline, vertical datum, units, target period, and uncertainty range. Finally, describe the historical value as an observed trend and the future value as a scenario-dependent projection. If the periods or definitions differ, explain the mismatch rather than averaging the values.
When to recalculate
Revisit a sea-level estimate whenever the underlying record or decision changes. Recalculate when a tide-gauge series receives additional observations, a station is relocated, an instrument or processing method changes, or a revised datum becomes available. Recheck the result after a new GNSS record changes the estimate of vertical land motion.
For future planning, update the analysis when a major assessment releases revised projection ranges, scenarios, baseline conventions, or downscaled coastal products. Also revisit it when the question changes—for example, from a regional trend to a specific property, bridge, wetland, or drainage system. A dataset appropriate for broad climate analysis may not have the spatial resolution or vertical reference needed for site-level decisions.
Use this final checklist before publishing or acting on a result:
- Name the measurement or projection type.
- State the location, period, units, and vertical reference.
- Separate relative sea-level change from ocean-surface change where possible.
- Include uncertainty and avoid unnecessary decimal places.
- Explain whether the number is observed, estimated, or projected.
- Save the dataset version, access date, and calculation steps.
That process makes a sea-level estimate easier to reproduce and easier to update. It also keeps the central distinction clear: global sea-level rise describes a broad ocean-system signal, while local risk depends on that signal plus land motion, tides, storms, coastal shape, and the elevation of the people and infrastructure at risk.
