How to choose a water quality intelligence platform (2026)
PEARL Intelligence Network · Reviewed
Choose a platform that can answer your water question with source-linked evidence for the place and period you care about. Before comparing feature totals, require a demonstration using your locations and records. The platform should make both the answer and its limitations understandable.
Start with the decision
A watershed assessment, an investigation of changing sensor readings, and a review of restoration performance need different evidence. Write down who will use the answer, what action it supports, the geography, and the acceptable age of the underlying data.
Then ask prospective providers to demonstrate the same task. A demonstration built around your question is more useful than a tour of unrelated dashboards.
Seven criteria to evaluate
1. The right evidence for the question
Ask which datasets support each output. EPA ATTAINS contains assessment and impairment information reported by states, territories and authorized tribes. EPA describes a two-year reporting cycle; that is different from a current sensor observation. USGS provides field and laboratory measurements, plus sensor observations at some monitoring locations. The Water Quality Portal provides discrete sample records from multiple organizations. These sources complement one another; they should retain their distinct meanings. EPA, USGS, Water Quality Portal
2. Traceability you can inspect
Choose a result and follow it to the publisher's record. Check the station or waterbody identifier, parameter, units, observation date, qualification flags and calculation method. A publisher logo alone does not establish where a particular number came from.
3. Separate observation time from retrieval time
A record downloaded today may describe conditions measured years ago. Require the platform to distinguish the observation date, the publisher's assessment or publication date where applicable, and the retrieval date. Ask what happens when a refresh fails: does the old value remain visibly dated, or appear current?
4. Geographic relevance
Verify the actual monitoring locations, waterbodies and jurisdictional boundaries used for your site. A station somewhere in the county does not necessarily represent your waterbody. A nearby utility does not, by proximity alone, establish a service relationship. Request the matching method and an explanation of coverage gaps.
5. Honest treatment of missing and uncertain data
Inspect a non-detect, an unavailable parameter, an old sample and a failed source request where those cases occur in the actual records. Missing data should remain missing. Ask whether the platform preserves detection limits, method changes and other qualifications when combining records or generating summaries.
6. Useful analysis with explicit limits
Ask the provider to distinguish observations, calculations, modeled estimates and recommendations. For an alert or score, request its inputs, evaluation evidence, applicability limits and review process. A statistical anomaly is a reason to investigate; it does not identify a contaminant or establish a regulatory determination by itself.
7. A workflow your organization can operate
Check role permissions, export contents, source links, revision history and support arrangements. Ask what remains readable outside the interactive dashboard. Obtain a written quote covering seats, implementation, refreshes, storage, support and exports rather than assuming the demonstration includes them all.
A practical selection exercise
- Select ten existing source records across three locations for a bounded demonstration.
- Include genuine stale, missing or qualified cases if they occur in that selection.
- Ask each provider to reproduce the same findings and export the underlying evidence.
Record each requirement as demonstrated, partly demonstrated or not demonstrated. Define mandatory requirements before viewing the results. An unmet source-provenance or access-control requirement should not disappear inside an average feature score.
Where PIN fits
PIN helps teams build the defensible record of what was known, when, and what was done about it. Before a site visit, report or complaint, PIN brings evidence, dates and the response into one reviewable record so teams can explain the decision and the action that followed.
PIN brings together 350+ federal and non-federal sources to support that accountability work. Use a demonstration to follow evidence for your locations from the publisher's record through the review and exported output. Ask who knew what, when it was available, what changed and how the response was recorded. PIN Water Intelligence
Coverage, freshness and available parameters vary by source and location. Source agencies retain their official records and determinations. PIN supports evidence review and accountability; it does not create a new laboratory result or replace field verification, professional judgment or an agency decision.
Frequently asked questions
Does national coverage mean my location has current measurements?
No. Confirm which parameters, locations and observation periods are actually available for your question.
Can a platform replace laboratory sampling or an agency decision?
Require clear boundaries. A public-record platform can support review, but a dashboard does not itself constitute a new laboratory result or an official agency determination.
Does a named data source guarantee current data?
No. Check the endpoint and version as well as the name. WQP's legacy interface warns that its WQX 2.2 profiles omit USGS data added after March 11, 2024; its newer profiles are linked separately. WQP notice