Guidance: Data visualisation

Purpose: Help partners turn monitoring results into clear visuals that support understanding, decisions, and action.

Who this is for: Catchment partnerships, volunteers, and delivery teams creating maps, charts, dashboards, and report cards.

When to use: After you have agreed your monitoring purpose and data management approach.

Principles and best practices

1. Start with the decision. Who needs to decide what, by when? Design the visual to answer that decision in one glance; deeper detail can sit behind tabs or tooltips.
2. Show thresholds and context. People need to see if a value is good, borderline, or poor. Use regulatory or locally agreed thresholds and make them visible (bands/lines/legends), not just implied.
3. Be honest about uncertainty. Add n-values, ranges, confidence bands, or icons for below-detection-limit. Explain caveats briefly near the graphic.
4. Favour comparisons over raw numbers. Trends, percent change, percentile ranks, and categories (e.g., traffic‑light) are easier to interpret than decimals.
5. Don’t over‑promise precision. Round sensibly; avoid implying precision greater than your method supports.
6. Keep it mobile‑first. Most viewers are on phones. Use short titles, large labels, clear legends, and single‑column layouts.
7. Design for equity and access. Use colour‑blind‑safe palettes, textures/markers as secondary encodings, high contrast, and plain language. Provide alt text (for screen readers).
8. Avoid acronyms and technical jargon. If you must use acronyms, spell them out or put in-context info buttons.
9. Document the method. Link to a short “How this was calculated” note: data source, date range, thresholds, and any weighting.
10. Minimise cognitive load. Each visual should communicate ONE idea. If you need more, split into small multiples.
11. Make updates obvious. Timestamp with “Last updated” and show the latest data point distinctly (e.g., an outlined marker).
12. Create space. Let your visualisations have adequate white space around them to allow viewers to visually group elements.

---

Choosing the right visual patterns

Use these patterns to match common CaSTCo questions:

1. Is this site improving or getting worse?

• Best: Time‑series line with threshold bands + quarterly/seasonal small multiples.
• Extras: Mark events (combined sewer overflow spills, heavy rain) as lollipops or shaded regions.

2. Where are the hotspots?

• Best: Choropleth/point map with traffic‑light categories; or a ranked bar chart of sites/sub‑catchments.
• Extras: Add a “Top 10 to investigate” list with links to site pages.

3. How does this compare to last year / other catchments?

• Best: Year‑on‑year small multiples, box/violin plots, or dumbbell charts for before/after.
• Extras: Show percent change badges (↑/↓ with magnitude).

4. Are we hitting agreed standards?

• Best: Threshold attainment chart (stacked bars showing % of samples in Good/Fair/Poor); or a gauge used sparingly.
• Extras: Include the standard name in the subtitle and link to method.

5. What’s the spatial pattern along a river?

• Best: Longitudinal profile (distance on x‑axis, parameter on y‑axis) with outfalls/tributaries annotated.
• Extras: Pair with a map overview for orientation.

6. How do multiple parameters look together?

• Best: Small multiples (one mini‑chart per parameter) or a tidy table with inline bars/sparks.
• Avoid: Dense radar charts… hard to read, rarely worth it.

---

Commonly used tools

Choose tools based on skill, governance, and longevity. Aim for portable outputs (PNG/SVG/PDF) and reproducible pipelines where possible.

Everyday office & design tools

• Google Sheets: Easy shared data entry, validation, and quick charts. Use data validation and consistent date/number formats; avoid merged cells; name ranges. Export CSV for platforms and PNG/SVG for charts.
• Microsoft Excel: Strong for pivot tables, conditional formatting, and one‑page “report card” layouts. Use defined styles and sparklines for trends; watch locale issues (decimal separators/dates). Export to PDF/PNG.
• Microsoft Word / Google Docs: Good for simple reports and printable one‑pagers. Use Styles for hierarchy, keep charts as linked images or pasted PNG/SVG, add alt text, set reading order, and export accessible PDFs.
• Canva: Fast layout with brand templates (great for report cards, posters, social). Canva charts are basic—ideally import charts from Sheets/Datawrapper as PNG/SVG. Check colour contrast and include alt text in captions.

No/low‑code charts & maps

• Datawrapper: Fast, accessible charts/maps; easy embeds; good for report cards and ranked lists.
• Flourish: Storytelling templates, scrollytelling, and simple maps.
• Looker Studio (Google) – Pull from sheets/BigQuery; simple dashboards for stakeholders.

GIS & web maps

• ArcGIS Online + Dashboards/Experience Builder – Enterprise dashboards, roles/permissions, field apps.
• QGIS (desktop) – Powerful open‑source GIS; great for producing static map outputs.
• Leaflet / MapLibre GL JS – Build bespoke, lightweight web maps; good when you have dev support.

Dashboards & BI

• Tableau / Power BI – Rich, interactive dashboards; good for organisations with licences.

Code and notebooks (reproducible)

• Python – pandas, matplotlib/plotly, geopandas, altair. Ideal for automated pipelines.
• R – tidyverse, ggplot2, sf, leaflet, quarto. Strong for statistical summaries and reporting.

Design & finishing

• Illustrator / Inkscape – Final polish for static figures.
• Figma – Team review, annotations, and asset handover.

Tip: Whatever you choose, save the source files and a frozen PNG/SVG/PDF with the same name and date (e.g., wr_nitrate_sitescore_2025-08-20.png).

Built‑in visualisation on monitoring platforms

Use these when you want quick, standard charts/maps with data governance and permissions already handled. They’re ideal for public sharing and stakeholder updates; step up to custom tools when you need unusual visuals or complex analysis.

• CaSTCo Data Hub
  • What you get: Directory of datasets with lightweight visualisations to browse results across partners.
  • Export/Embed: Dataset‑level downloads or links back to the source platform.
  • API/Integration: Works as an aggregator; underlying sources may expose their own APIs.
• CaBA portals
  • What you get: ArcGIS‑based web maps/dashboards with standard layers and indicators.
  • Export/Embed: Layer downloads/embeds depend on the hosting organisation’s settings.
  • API/Integration: Many layers have ArcGIS REST endpoints; availability varies by layer.
• FreshWater Watch (Earthwatch)
  • What you get: Site maps, time‑series by parameter, event summaries.
  • Export/Embed: CSV downloads; public pages for sharing.
  • API/Integration: In some deployments, partner access or feeds are available—check your project setup.
• Cartographer
  • What you get: Public map of surveys, hotspot/scorecard views, set up for Riverfly.
  • Export/Embed: CSV/GeoJSON and embeddable views (varies by instance).
  • API/Integration: JSON endpoints may be available for selected resources; confirm with the host.
• Water Rangers
  • What you get: Site pages with time‑series and longitudinal views; printable summaries, customisable data forms.
  • Export/Embed: CSV/GeoJSON per site/dataset; shareable links.
  • API/Integration: Documented read‑only API endpoints for sites, observations, and parameters

Tips

• Prefer platform exports (CSV/GeoJSON) as the single source of truth; avoid manual copying.
• If using APIs, cache data and display a “Data through: YYYY‑MM‑DD” note to manage expectations.
• Keep a small data dictionary mapping platform field names → your report card fields.

Workflow, versioning, and governance

• Pipeline: Data entry → QA/QC → aggregation → visualisation → review → publish.
• Roles: Name an owner for each visual/report card. Use a simple approval step before publishing.
• Versioning: Store data snapshots and code/notebooks; keep a changelog of rules/thresholds.
• Timestamps: Display “Data through: YYYY‑MM‑DD” and “Last updated: YYYY‑MM‑DD”.
• Licensing & credit: Add a short attribution and licence (e.g., ODbL/CC‑BY) where applicable.

Data visualisation inspiration

Here are just a few examples of data visualisation examples. More resources and templates for data visualisation are under development.