A published case study reports a number. Whether that number was measured on the ground, modelled for a design storm or estimated from a coefficient decides how far it can be carried to the next project. The same brief often mixes all three, and the headline figure may be the weakest one.
We read six projects in different climates, mostly through the Landscape Architecture Foundation’s Landscape Performance Series, with the monitoring reports and papers behind them. For each we record what was reported, the basis, who produced it, and the period. This is a reading guide, not engineering advice, and it does not rank the projects.
How the Landscape Performance Series frames evidence
The Foundation defines landscape performance as how well a landscape solution meets its intended purpose, judged by measurable progress toward environmental, social and economic goals. LAF, About Landscape Performance Its guide to evaluating performance says it presents possibilities and not a prescriptive list. Its four criteria for metrics and methods are: easy for non-experts to use, applicable across project types and scales, measurable in a short time with a limited budget, and defensible. LAF, Guide to Evaluate Performance
Our interpretation: a method chosen to fit a short time and a limited budget will often be a calculation or a short sampling effort, so each benefit needs checking against its own methods document. The Middle Blue River methods document says the programme matches faculty-student teams with practitioners to document exemplary, high-performing projects. LAF, Middle Blue River Methods The projects were chosen as good examples, so the set says little about how often such projects underperform.
Seattle: a monitored street, with a short baseline
Seattle Public Utilities rebuilt 2nd Avenue NW as a Street Edge Alternatives (SEA Streets) project, with reduced impervious area and detention ponds. University of Washington researchers monitored flow against rainfall. At 2nd Avenue NW, flow was sensed with a float and shaft encoder behind a V-notch weir at the point where runoff leaves the project, and rain with a tipping bucket gauge. The report is dated November 2002. Horner, Lim and Burges 2002
The study has a pre-construction period, March to July 2000, with 35 storms, and all 35 produced flow over the weir. After construction, only 10 of 96 rain events produced discharge, measured or estimated. Over January 20, 2001 to April 30, 2002, 98.2% of the rainfall volume on the catchment was not discharged, with 97.6% for the 2001–2002 wet season and 100% for the 2001 dry season. Horner et al. 2002
Part of that record is inferred. The report says undetected slippage of the shaft encoder made its readings unreliable from March 2001 to April 2002, and discharge for parts of that period came from manual stage measurements. For the remaining gaps, the authors assumed no discharge for events with less than 2,300 ft³ of rainfall volume, because no discharge had ever been measured below that volume. Seven larger unmeasured events were estimated at 5% of precipitation volume, based on three measured storms that ranged from 3.1% to 6.8%. The authors put the likely error at no more than about 2%. Horner et al. 2002
The report attributes flow reduction primarily to infiltration and describes the volume decreases as indicative of recharge. Infiltration was not measured. It is inferred from rainfall on the catchment and discharge at the single exit weir. The report monitored flow and rainfall only. It states water quality benefits as expectations and reports no water quality sampling. Horner et al. 2002
Four points from the report limit the figure. Horner et al. 2002
- The rainfall varied. The report lists the 2000–2001 wet season (October to March, at the airport station) at 57% of the 52-year mean and 2001–2002 at 109%. Neither winter had any very large storms.
- The comparison with a conventional street is an estimate. The report puts wet-month discharge at a factor of 4.7 lower than for a street designed to City of Seattle conventions. That street was not built or measured.
- The baseline is short. The report calls its comparison with the original street only indicative, because the baseline period was abbreviated and skewed toward the drier months.
- The system’s capacity is stated. It fully attenuates about 2,300 ft³, the volume from about 0.75 inch of rain on the catchment.
The owner’s project page adds design context. It says the original hope of infiltrating along the whole block was not possible because some homes had a groundwater intrusion problem, so some swales were given an impermeable liner. It says the swales were sized to detain the pre-developed two-year, 24-hour storm while not allowing infiltration. It also says two years of monitoring show a 99% reduction in the volume leaving the street. The page gives no method, source or date for that figure, so we treat it as the owner’s summary. Seattle Public Utilities, Street Edge Alternatives Project
Kansas City: measured and modelled results in one brief
The Middle Blue River Basin Green Solutions Pilot Project in Kansas City, Missouri, was completed in 2012. Its brief lists a runoff reduction of about 80% and a peak flow reduction of 76%. LAF brief The brief does not say how each was obtained. The methods document does. LAF methods
- Peak flow 76% (9.2 cfs) and volume 36% for a 1.4-inch storm: modelled. The methods document, citing the City’s 2013 final report, says reliable post-construction flow data was not available, so a model was used. URS, the project’s prime design consultant, ran it in XPSWMM. The model was calibrated to Darcy’s law equations and then to three 2013 rain events and one flow meter’s data from April 2013. Only that one meter’s data was available at calibration. LAF methods
- Runoff about 80%: measured, with corrections. A University of Missouri-Kansas City graduate student, funded by EPA through Tetra Tech, measured the 100-acre pilot area and an 80-acre control with four flow meters. Pre-construction monitoring covered 24 rain events from December 2008 to June 2010, then 6 events from February 2011 to May 2012, a period in which rain garden construction started. The post-construction period covered 8 events, from November 2012 to May 2013. During the pipe repair period the meters recorded velocity only, so flow was reconstructed. Runoff percentage changed from 21% before the pipe repair to about 41% after it, which the document reads as leakage that had under-measured earlier runoff. It was 7.6% after construction, giving about 81.5%. LAF methods
- 360,320 gallons per storm: design capacity. The document states this is what is available, not measured use. LAF methods
- Tree carbon and canopy interception: estimated with the National Tree Benefit Calculator from an inventory. LAF methods
Our reading: the 80% figure has a real measurement behind it, from eight storms in one winter and spring, with a corrected baseline. It is evidence that the rain gardens reduced runoff. It is not a long-term record.
Los Angeles: a calculation, some sampling and a survey
The Elmer Avenue Neighborhood Retrofit in the Sun Valley area of Los Angeles was completed in 2010, with the case study prepared in 2011. Its brief reports that the project infiltrates up to 5.4 million gallons a year. It notes a Council for Watershed Health monitoring programme. LAF brief
The methodology shows that the 5.4 million gallons is a calculation: 16 inches of average annual rainfall, a 0.31 runoff coefficient and a 40-acre area, which gives 5,387,000 gallons. It is the runoff the area is expected to generate. It is not a metered infiltration volume. The 60%, 33% and 18% reductions in lead, copper and total suspended solids come from water sampling reported in a 2011 StormCon paper, for water passing through a catch basin sump. The 30% and 10% potable water savings rest on a landscape architect’s statement of Metropolitan Water District research, and the document words the saving as an expectation. The carbon estimate rests on an agronomist’s assessment of potential and is expected to fall from 7.25 tons to about 2 tons by year 8. The walkability survey had 13 responses from 24 households. LAF methodology
The Council’s 2015 final report to the Bureau of Reclamation on the neighbouring Elmer Paseo adds later information. It says monitoring equipment had been deployed at the Paseo but that, because of the drought, there was not yet enough data to express actual flow and capture values. It also says monitoring funded by that grant allowed improvements to the Elmer Avenue project that doubled its infiltration capacity. Council for Watershed Health 2015 The report gives the two capacity figures without a unit, so we do not quote them. Our interpretation is that the 2011 figure describes the project as first built.
Iowa: design-storm modelling, labelled as such
Charles City Permeable Streetscape Phase 1 was completed in 2009, and the case study was prepared in August 2012. The brief reports peak flow reductions of at least 75% for 10-year storms and 40% for 100-year storms. LAF brief The methodology states how: a HydroCAD model using the NRCS curve number method and the unit hydrograph method, soil group B, and rainfall data from a Midwest Climate Center report. It credits the results to the 2009 Charles City Green Streets design report and labels them as the proposed condition. It notes exceptions to the headline figures on Johnson and, for the 100-year storm, at Iowa and Ferguson. The infiltration inputs were design values, and for the sandy soil beneath the pavement the model used a conservative 0.88 inch per hour against a tested average of 1.4. LAF methodology
The volume result is a prototype model of one block, Hulin Street between Joslin and Iowa Streets. It shows zero discharge up to the two-year storm and more than a 60% volume reduction for the 10-year storm. The salt reduction of up to 75% is described as expected, from University of New Hampshire Stormwater Center research, and not verified at this site. LAF methodology The brief adds that the cobble infiltration areas of Phase 1 collected debris, lost effectiveness and were dropped in Phase 2. LAF brief Our reading is that the model, built on design infiltration values, did not represent that maintenance loss.
Phoenix: a short heat survey
Phoenix Civic Space Park, completed in 2009 in a hot desert climate, reports air temperatures 1.8°F lower than a typical urban landscape and mid-day surface temperatures 12.4°F lower in turf areas and 23.4°F lower in hardscape. LAF brief
The methodology, from an Arizona State University team in 2014, gives the basis. LAF methodology
- Surface temperature: a hand-held infrared thermometer at 72 points on a 50-foot grid, at solar noon and again at night, on one day, June 5, 2014.
- The surface reductions: these compare shade with open sun inside the park on that day. For concrete the table shows 125.7°F in the open and 102.3°F in shade. For turf it shows 88.3°F and 75.9°F. They do not compare the park with its surroundings.
- Air temperature: data loggers recording every 30 minutes for three days, June 5 to 7, 2014. Twelve loggers sat under tree canopy. Two of the twelve were in similar trees on a nearby streetscape as the comparison. One more sat in the open in a weather shield. The park averaged 91.7°F and the streetscape 93.5°F.
- Night effects: shaded surfaces cooled less in the evening than open ones, and paved areas in the park had the highest air temperatures around 4 to 6 am. The authors suggest canopy may block the release of long-wave radiation. That was not tested.
Our reading: the figures describe three clear, hot days in one June, two comparison loggers and one day of surface readings. They show the direction and rough size of shade effects. They give no annual or heat-wave result, and they do not show how much of the 1.8°F difference comes from the park’s design.
Louisiana: a coastal measurement that found no protection
La Peyre and colleagues examined experimental oyster reefs built in 2009 in a Louisiana estuarine lake as a nature-based shoreline defence. They compared six reefs with paired reference sites in 2009–2011 and again in 2019–2020. Shoreline erosion was about 1 metre per year at both reef and reference sites, immediately after construction and 10 years later, which the authors read as a lack of shoreline protection. Oyster density fell from more than 1,000 per square metre to below 60. The authors say this likely reflects missing new recruits after repeated low-salinity periods, including a record low in 2019. They note that reproductive-sized oysters persisted and that the reefs could still develop benefits if conditions improve. La Peyre et al. 2022, USGS record
This is an experimental reef study, not a landscape architecture case study. It is included because it has reference sites and two time points ten years apart, and because it reports a null result. We read the USGS record and abstract, not the full paper.
How the six compare
| Project | Headline result | Basis | Period |
|---|---|---|---|
| Seattle SEA Streets | 98.2% of rainfall volume not discharged | Exit weir and manual readings; some events estimated or assumed zero | About 15 months |
| Kansas City | About 80% runoff reduction | Measured, with corrections | 8 post-construction events |
| Kansas City | 76% peak flow | Modelled | One design storm |
| Los Angeles, Elmer Avenue | 5.4 million gallons a year | Calculated from rainfall, coefficient and area | Not applicable |
| Charles City | 75% and 40% peak flow | Modelled, proposed condition | Design storms |
| Phoenix | 1.8°F, 12.4°F, 23.4°F | Measured | 3 days air, 1 day surface |
| Louisiana oyster reefs | No protection from erosion | Measured against reference sites | 2009–2011 and 2019–2020 |
The table is our summary. The Seattle period is our count from the report’s dates. The other periods are as the sources state them.
How much weight, and for what
This is interpretation. A case study number is evidence for what a type of design can do. It does not predict a new site.
- Find the basis first. The methods document, the monitoring report or the paper decides it. The briefs did not always state it.
- Look for a baseline, and check its length. Kansas City’s measured figure depends on one. Seattle’s four-month baseline supports only an indicative comparison. Phoenix used two loggers on one nearby streetscape as a stand-in.
- Check the period against the climate. The two Seattle wet seasons ran from 57% to 109% of mean rainfall. Kansas City covers eight storms.
- Ask what was inferred inside a measured figure. The Seattle result includes assumed zero discharges and estimated events. Part of the Kansas City baseline was reconstructed.
- Read capacity figures as capacity. They describe what the design can hold.
- Note what was not measured. Seattle has no water quality sampling and no direct measurement of infiltration. Maintenance was not in the Charles City model.
- Check who measured and who paid. University researchers monitored Seattle under a memorandum of understanding with the owner. Horner et al. 2002 EPA funded the Kansas City measurement through a contractor. The Elmer Avenue water-use and carbon figures rest on advice from the project team.
Using this in Field Method
A later comparison needs a dated record of conditions before construction. Site Dossier assembles a source-linked public-data record for a point in San Diego County, with each layer shown with its source and date. The page describes it as mapped, not surveyed. That record can serve as a starting baseline. It does not monitor performance. Our related reading covers how far to trust stormwater and microclimate simulation tools, site-scale strategies for heat, stormwater and drought and what rating-system levels do and do not show.
What we could not settle
- How much of the Seattle volume infiltrated and how much left by evaporation and plant uptake. The report infers infiltration and did not measure it.
- What later Seattle monitoring shows. The owner’s page cites two years of monitoring without a source, and we did not open a later report.
- What the 2011 StormCon paper reports for Elmer Avenue, and what monitored infiltration there has been since the improvements.
- What the Kansas City final report and the University of Missouri-Kansas City thesis show beyond the LAF methods document.
- How often Landscape Performance Series briefs report a failed or reduced benefit. We did not survey the series.
- Measured heat results over a full summer or heat wave for a designed park.
- Measured flood reduction at the neighbourhood or watershed scale.
Sources
- Landscape Architecture Foundation — About Landscape Performance
- Landscape Architecture Foundation — Guide to Evaluate Performance
- Landscape Architecture Foundation — Middle Blue River Basin Green Solutions Pilot Project, Methods Document for Landscape Performance Benefits (2016)
- Horner, Lim and Burges (2002) — Hydrologic Monitoring of the Seattle Ultra-Urban Stormwater Management Projects, Water Resources Series Technical Report No. 170
- Seattle Public Utilities — Street Edge Alternatives Project
- Landscape Architecture Foundation — Middle Blue River Basin Green Solutions Pilot Project (case study brief)
- Landscape Architecture Foundation — Elmer Avenue Neighborhood Retrofit (case study brief)
- Landscape Architecture Foundation — Elmer Avenue Neighborhood Retrofit, Methodology for Landscape Performance Benefits
- Council for Watershed Health — The Elmer Paseo Stormwater Improvements Project, Final Report to the US Bureau of Reclamation (2015)
- Landscape Architecture Foundation — Charles City Permeable Streetscape Phase 1 (case study brief)
- Landscape Architecture Foundation — Charles City Permeable Streetscape Phase 1, Methodology for Landscape Performance Benefits (2012)
- Landscape Architecture Foundation — Phoenix Civic Space Park (case study brief)
- Landscape Architecture Foundation — Phoenix Civic Space Park, Methodology for Landscape Performance Benefits (2014)
- La Peyre, Buie, Rossi and Roberts (2022) — Long-term assessments are critical to determining persistence and shoreline protection from oyster reef nature-based coastal defenses (USGS publication record)
Try a relevant tool
These tools support site review and design studies. Read each tool’s stated capabilities and limits before use.
Explore Site Dossier