OverviewDataset Summary
Every Clean Water Act assessment unit in Northern Virginia from the EPA ATTAINS 2022 reporting cycle: 344 stream and river line units totaling 4,828.8 assessed km, plus 56 lake, reservoir and estuarine polygons covering 32.6 sq km. Each unit carries its integrated report category (2 through 5), impairment and 303(d) status, TMDL progress, designated-use support for recreation, aquatic life, fish consumption and drinking water, and a synthesized list of impairment causes. 221 units are impaired - 919.2 stream km and 28.4 sq km of open water. The map colors each water by status: red for Category 5 (303(d) list), orange for impaired with an approved TMDL (4A), gray for insufficient data, and green for fully supporting waters.
Why This DatasetThe Problem It Solves
Impaired-waters data governs real money: MS4 stormwater permits, TMDL waste load allocations, stream restoration crediting and development stormwater requirements all key off the 303(d) list. Yet the authoritative data lives in EPA's ATTAINS system as a national service with no regional clip, no jurisdiction totals, and impairment causes scattered across roughly sixty per-parameter columns. This atlas reduces that to a clean NoVA product: one line layer, one area layer, a per-unit cause list, and jurisdiction summaries that show at a glance that one in five assessed stream kilometers in the region fails a designated use - and that in the urban core (Arlington, Fairfax City) it is one in two.
MethodologyHow It Was Built
EPA ATTAINS Assessment MapServer layers 1 (lines) and 2 (areas) were queried by ArcGIS REST GeoJSON export with the NoVA bounding envelope and a state=VA filter (462 raw lines, 61 raw areas, 2022 reporting cycle). Features were clipped to the union of the 9 NoVA jurisdictions (TIGER cb_2023 cartographic boundaries; FIPS 51013, 51059, 51107, 51153, 51510, 51600, 51610, 51683, 51685), yielding 344 line and 56 area assessment units after deduplication to the most recent cycle per unit. Jurisdiction was assigned by representative-point spatial join. The ~60 per-parameter cause columns (values Cause/Meets) were collapsed into a semicolon-delimited cause list per unit. Clipped lengths were computed in EPSG:32618 (4,828.8 km total); areas in sq km. Outputs: two GeoPackages with embedded FGDC metadata and XML sidecars, a flat CSV, and web GeoJSONs simplified to 0.0002 degrees. DBSCAN clustering, anomaly detection and composite indexing were not run: the geometry is a hydrographic line network (not a point layer), the extract is a single 2022-cycle snapshot with no per-feature time series, and IR category plus designated-use columns already provide the categorical scoring an index would duplicate.
Key FindingsWhat the Data Shows
- One in five assessed stream kilometers is impaired: 919.2 of 4,828.8 assessed km (19.0%) fail at least one designated use; 221 of 400 assessment units (55%) are impaired, and 46 of 56 lake/estuary polygons (28.4 of 32.6 sq km) carry an impairment.
- Pathogens are the region's dominant impairment: 153 units (38%) list pathogens (E. coli) as a cause - the classic urban/suburban stormwater signature - and recreation use is Not Supporting on exactly those 153 units. Benthic community impairment (cause unknown, impaired biota) follows with 88 units, then PCBs with 64.
- The 303(d) backlog is half-treated: 115 units sit in Category 5 (the formal 303(d) list needing a TMDL) while 106 units have progressed to Category 4A (TMDL approved); 183 units overall have an approved TMDL - evidence of two decades of bacteria and benthic TMDL work in the Occoquan, Accotink, Four Mile Run and Bull Run watersheds.
- Urban cores are proportionally worse: Arlington County (50.4%) and Fairfax City (55.4%) have the highest impaired shares of assessed length, while exurban Prince William sits at 15.0% - impairment tracks imperviousness, not jurisdiction size.
- The longest impaired streams are exurban growth corridors: Little Bull Run (14.4 km), Horsepen Run (13.2 km), Four Mile Run (12.8 km), Cedar Run (12.6 km) and Kettle Run (12.5 km) - three of the five drain Prince William's fastest-developing land.
- The Occoquan Reservoir - a drinking water source for ~800,000 people - is impaired: the 5.06 sq km reservoir polygon, Gunston Cove (3.9 sq km) and the tidal Potomac mainstem all carry impairments including PCBs in fish tissue.
Analytical UsesWhat You Can Do With This
- MS4 permit compliance: every NoVA jurisdiction holds an MS4 stormwater permit with TMDL-driven pollutant reduction requirements; this layer maps exactly which receiving waters drive those obligations.
- Development due diligence: projects draining to a 303(d)-listed or TMDL water face additional stormwater and E&S requirements; a parcel-level intersect flags exposure before land acquisition.
- Stream restoration targeting: benthic-impaired segments (88 units) are the candidate pool for county stream restoration credits under the Chesapeake Bay TMDL.
- Drinking water source protection: impairments upstream of the Occoquan Reservoir intake support source-water protection planning for Fairfax Water.
- Equity and recreation analysis: 153 units not supporting recreation use can be joined to demographics to ask who lives near swimmable versus contact-restricted water.
Synthesis OpportunitiesWhat It Combines With
Intersect with the Day 10 stream network to extend impairment attribution to unassessed headwaters by catchment. Join the Day 39 traffic atlas to test road-salt and highway-runoff corridors against chloride and benthic impairments. Overlay the Day 33 data center atlas to screen cooling and construction discharge exposure in the Broad Run and Bull Run watersheds. Combine with the Day 29 USGS gage network to pair impairment status with measured discharge trends, and with the Day 2 flood risk index to find tracts where flooding and contaminated water co-occur.
LimitationsKnown Caveats
- Assessed waters only: Virginia DEQ assesses a monitored subset; unassessed headwaters (Category 3 covers 129 units with insufficient data) are not clean, just unmeasured.
- 2022 reporting cycle: the ATTAINS national geospatial service publishes the 2022 integrated report vintage; Virginia's 2024 cycle exists in tabular ATTAINS but is not yet in this map service.
- Clipping effects: border streams (Potomac mainstem, Bull Run) are clipped at the jurisdiction union; lengths are NoVA-side only and not comparable to DEQ statewide totals.
- Cause granularity: cause flags are unit-level, not reach-level; a long unit may be impaired only in part.
- Two jurisdictions report zero units: Falls Church City and Manassas Park City contain no assessment-unit representative points - their small streams are assessed under units attributed to neighboring counties.
Data DictionaryFields in the GeoPackage
| Field | Type | Description | Units | Null OK? |
|---|---|---|---|---|
| assessmentunitidentifier | Text | Virginia DEQ assessment unit ID | - | No |
| assessmentunitname | Text | Water body name | - | No |
| ircategory | Text | Integrated report category (2, 3, 4A, 5) | - | No |
| overallstatus | Text | Overall status (Impaired / Fully Supporting / Not Assessed) | - | No |
| isimpaired / isthreatened | Text | Impairment flags (Y/N) | - | No |
| on303dlist | Text | On the 303(d) list requiring a TMDL (Y/N) | - | No |
| hastmdl | Text | Has an approved TMDL (Y/N) | - | No |
| recreation_use / ecological_use / fishconsumption_use / drinkingwater_use | Text | Designated-use support status | - | Yes |
| causes | Text | Semicolon list of impairment causes | - | Yes |
| n_causes | Integer | Number of impairment causes | count | No |
| jurisdiction | Text | NoVA jurisdiction (representative point) | - | Yes |
| length_km | Real | Clipped line length, EPSG:32618 (lines GPKG) | km | No |
| area_sqkm | Real | Clipped polygon area (areas GPKG) | sq km | No |
| waterbodyreportlink | Text | EPA How's My Waterway report URL | - | Yes |
Jurisdiction BreakdownResults by Jurisdiction
| Jurisdiction | Line Units | Assessed km | Impaired Units | Impaired km | % km Impaired |
|---|---|---|---|---|---|
| Loudoun County | 119 | 1,944.6 | 73 | 363.1 | 18.7% |
| Fairfax County | 145 | 1,374.6 | 80 | 314.1 | 22.9% |
| Prince William County | 116 | 1,407.4 | 58 | 211.8 | 15.0% |
| Arlington County | 7 | 31.8 | 2 | 16.0 | 50.4% |
| Alexandria City | 9 | 52.4 | 7 | 8.9 | 16.9% |
| Fairfax City | 2 | 9.7 | 1 | 5.4 | 55.4% |
| Manassas City | 2 | 8.3 | 0 | 0.0 | 0.0% |
| Falls Church City | 0 | 0.0 | 0 | 0.0 | - |
| Manassas Park City | 0 | 0.0 | 0 | 0.0 | - |
Sources & CitationsAll Data is Openly Licensed
- [1]U.S. EPA, ATTAINS Assessment geospatial service, 2022 integrated reporting cycle, gispub.epa.gov, accessed July 3, 2026.
- [2]Virginia DEQ, 2022 305(b)/303(d) Integrated Report.
- [3]U.S. Census Bureau, TIGER/Line cartographic boundaries cb_2023.
- [4]Keystone GIS derived products: NoVA clip, cause synthesis, jurisdiction stats. CC BY 4.0.