---
title: "Flood Risk at Lighthaven"
author: GPT-6 Astra, Claude-5.1-Fable, Gemini-4 Argon, Gwern
description: "Lighthaven, Berkeley, sits about 64 meters above sea level in FEMA Zone X, but its history includes rain-related flooding. Municipal records identify nearby drainage constraints; present building-level risk depends on repairs, surface drainage, and pump systems."
created: 2026-10-28
modified: 2026-10-28
status: finished
confidence: highly likely
...
Lighthaven is a conference venue with overnight lodging in Berkeley, California, occupying the former Rose Garden Inn at 2740 and 2744 Telegraph Avenue and 2348 Ward Street.
Its sampled ground elevations are **63.0–65.2 m (206.7–213.8 ft)** above the North American Vertical Datum of 1988, and it lies in **unshaded FEMA Zone X**, outside the mapped 1% and 0.2% annual-chance floodplains.
The reviewed tsunami, coastal-inundation, and Claremont Dam breach maps do not place the campus in an inundation zone.
Elevation data, municipal geographic information, engineering reports, and accounts of the property's operation nevertheless establish a local rainwater concern.
In December 2005, the Rose Garden Inn's manager acknowledged rain-related flooding in the Fay House.
Berkeley subsequently identified Telegraph Avenue at Stuart Street as a localized-flooding location, and its 2011 drainage model predicted overflow at that intersection.
The principal unresolved exposure is rainfall interacting with street drainage, courtyard grading, lower-level entrances, and any groundwater or pumping systems.
The evidence supports low exposure to the mapped large-area flood hazards, but it does not establish a numerical probability, depth, or cost of flooding inside today's buildings.
The decisive missing information concerns repairs since the documented incident and the current paths by which water enters, leaves, or is pumped from the property.
**Scope and confidence.**
This assessment concerns the existing Lighthaven campus and immediately adjoining drainage, with evidence checked on **2026-10-28**.
The central assessment is **highly likely**; present-day building exposure remains unquantified.
This is a synthesis of public records and terrain data; it does not supply a surveyed finished-floor elevation or a calibrated model of water movement through the campus.
| Question | Assessment | Evidence boundary |
|:--|:--|:--|
| How high is Lighthaven? | About **64 m (210 ft)**; sampled terrain spans **63.0–65.2 m** | Ground-model estimates, not floor or doorway surveys |
| Has the property flooded? | **Yes: rain-related flooding was acknowledged in December 2005** | Entry route, depth, repair success, and later recurrence are not established |
| What is its FEMA zone? | **Unshaded Zone X**, minimal mapped hazard | Does not quantify all local rainwater or drainage failures |
| Is direct Bay inundation a concern? | No exposure in the reviewed coastal scenarios | Separate from rain-fed groundwater and local runoff |
| Is tsunami inundation mapped here? | No | Applies to the published planning zone |
| Does the reviewed dam-breach map reach it? | No: the Claremont footprint passes south of the campus | Specific modeled breach scenarios, not a guarantee about every infrastructure failure |
| What warrants attention? | Rainwater entry, surface runoff, and drainage performance | Current severity depends on site conditions and repairs |
Table: Lighthaven combines favorable regional flood mapping with a documented history of local rain-related flooding; the evidence and limitations are developed below.
**Site and terrain.**
The campus occupies three contiguous parcels on the west side of Telegraph Avenue, between **Ward Street to the north and Stuart Street to the south**.
The addresses are 2740 Telegraph Avenue, 2744 Telegraph Avenue, and 2348 Ward Street.
Berkeley's planning record identifies the three-parcel, five-building former inn and provides a site plan [Berkeley 2020](https://berkeleyca.gov/sites/default/files/legislative-body-meeting-agendas/2020-02-5%20PC%20Agenda%20PACKET.pdf#page=37 "Rose Garden Inn rezoning materials, Planning Commission packet, 5 February 2020").
Derby Street lies north of Ward Street.
[Terrain estimates.]{.marginnote}
Twenty-two point queries across the campus returned elevations from 63.0 to 65.2 m.
A representative central point, at 37.85991° N, 122.25966° W, returned 64.7 m (212.3 ft).
The United States Geological Survey (**USGS**) service identifies the underlying one-meter terrain product as `CA_AlamedaCounty_2021_B21` and its vertical datum as the **North American Vertical Datum of 1988 (NAVD88)** [USGS n.d.](https://elevation.nationalmap.gov/arcgis/rest/services/3DEPElevation/ImageServer/89924?f=pjson "USGS 3DEP elevation raster metadata: CA_AlamedaCounty_2021_B21").[^elevation]
[^elevation]: The reproducible [central-point query](https://epqs.nationalmap.gov/v1/json?x=-122.25966&y=37.85991&units=Feet&wkid=4326&includeDate=true "USGS elevation query at the representative Lighthaven central point") requests feet in longitude/latitude coordinates; conversions use exactly 0.3048 m per foot.
The 22 samples include boundary and interior points, not an exhaustive search for every parcel maximum or minimum.
The slope estimate fits elevation to local east and north distances; its gradient magnitude is approximately 0.0283.
The product's name contains 2021, but the service metadata does not justify assigning every terrain measurement one precise acquisition day.
| Sample location | Latitude | Longitude | Elevation, m | Elevation, ft |
|:--|--:|--:|--:|--:|
| Representative center | 37.8599100 | −122.2596600 | 64.7 | 212.3 |
| Eastern interior sample | 37.8599000 | −122.2595000 | 65.2 | 213.8 |
| Western sample | 37.8598910 | −122.2599693 | 63.5 | 208.2 |
| Southwestern sample | 37.8595298 | −122.2598966 | 63.0 | 206.7 |
Table: Selected USGS elevation queries show the broad westward descent; displayed precision describes the returned estimates, not surveyed accuracy.
A plane fitted to the 22 samples has a slope of approximately 2.8%, descending predominantly westward, with a smaller southward component.
This calculation summarizes the terrain; it does not resolve individual steps, walls, gutters, planting beds, or drain inlets.
USGS cautions that point-query elevations are interpolated from digital elevation models and are not precisely measured survey elevations [USGS n.d.](https://www.usgs.gov/faqs/how-accurate-are-elevations-generated-elevation-point-query-service-national-map "How accurate are elevations generated by the Elevation Point Query Service in The National Map?").
For regional inundation, the difference between 63 and 65 m is immaterial.
For water entering a particular room, the difference between a threshold and the adjacent courtyard may be decisive.
**FEMA flood mapping.**
The **Federal Emergency Management Agency (FEMA)** maps the campus as **unshaded Zone X**, an area of minimal mapped flood hazard outside the mapped 1% and 0.2% annual-chance floodplains.
A query of Berkeley's flood-hazard layer over an envelope enclosing the campus returns `FLOODZONE = X`; the layer separately identifies the 0.2%-annual-chance category.
The City's older layer also marks the location as outside the **Special Flood Hazard Area (SFHA)** [Berkeley n.d.](https://gis.cityofberkeley.info/arcgis3/rest/services/Public/Portal_Environment/MapServer/3 "City of Berkeley flood-hazard GIS layer").[^map-query]
[^map-query]: The [campus-envelope query](https://gis.cityofberkeley.info/arcgis3/rest/services/Public/Portal_Environment/MapServer/3/query?f=json&geometry=-122.26005%2C37.85948%2C-122.25925%2C37.86035&geometryType=esriGeometryEnvelope&inSR=4326&outSR=4326&outFields=%2A&returnGeometry=false "Berkeley flood-hazard query over the Lighthaven campus envelope") covers longitude −122.26005 to −122.25925 and latitude 37.85948 to 37.86035.
It returns the X polygon, object 950.
A separate check of Esri's [2026-03-03 NFHL snapshot](https://www.arcgis.com/home/item.html?id=ccd29fb4bd5e4b44a2d61638829a5d7e "USA Flood Hazard Areas: Esri view of FEMA National Flood Hazard Layer data") finds no displayed hazard polygon there.
That view omits both minimal-hazard and some unknown/no-data categories, so its empty result corroborates the positive City classification without independently proving it.
These are query and snapshot dates, not FIRM effective dates.
The relevant **Flood Insurance Rate Map (FIRM)** panel is **06001C0057G**.
A survey in the City's nearby 2712–2714 Telegraph Avenue records identifies that panel and its effective date, **2009-08-03** [Berkeley 2026](https://berkeleyca.gov/sites/default/files/2026-04/2026-04-09_ZAB_Item%206d_2712%20Telegraph_Staff%20Report%20and%20Attachments.pdf "2712 Telegraph Avenue staff report and attachments, including ALTA survey, 9 April 2026").
The neighboring survey supports the panel identification; the campus-wide GIS query supplies the site-specific classification.
A later county flood-study revision, an updated GIS service, or a recent retrieval date should not be mistaken for a new effective date for this individual panel.
[What Zone X means.]{.marginnote}
Zone X does not assign the probability of every way that water can enter a building.
FEMA expressly recognizes that buildings outside the SFHA can flood when severe rainfall combines with inadequate local drainage [FEMA n.d.](https://agents.floodsmart.gov/sites/default/files/bulletins/W-02003/w-02003.pdf "National Flood Insurance Program bulletin explaining Zone X and local drainage flooding").
Consequently, the mapped exclusion cannot be converted into a claim that Lighthaven has less than a 0.2% annual chance of *any* flood damage.
The distinction between a mapped creek floodplain and an overwhelmed street drain is central to this site.
Zone X does not ordinarily trigger the federal SFHA mandatory-purchase requirement for flood insurance, but a lender may still require coverage [FEMA n.d.](https://www.fema.gov/sites/default/files/2020-08/fema_commonly-used-terms-acronyms-levee-system.pdf "Commonly Used Terms and Acronyms: lender discretion outside the Special Flood Hazard Area").
Nor does the map alone establish exemption from every local development rule: Berkeley Municipal Code §17.12.060 calls the FEMA mapping the minimum area of applicability and allows supplemental studies [Berkeley 2009](https://berkeley.municipal.codes/BMC/17.12.060 "Basis for establishing the areas of special flood hazard").
**Coastal, tsunami, and dam hazards.**
At approximately 64 m elevation, Lighthaven is well above the coastal water levels considered in Berkeley's published inundation planning.
Direct Bay storm surge, king tides, and the reviewed sea-level-rise scenarios do not identify a campus inundation hazard.
Berkeley's tsunami zone remains near the waterfront, extending roughly to Third through Seventh Streets depending on location, well west of Lighthaven [Berkeley 2025a](https://berkeleyca.gov/sites/default/files/2026-06/Berkeley%20Hazard%20of%20Concern-%20Accessible%20Text%20Version.pdf#page=13 "Berkeley’s Hazards: accessible text of the 2024 hazard analysis, August 2025").
The **East Bay Municipal Utility District (EBMUD)** Claremont Dam map, issued in June 2021 from a January 2021 breach analysis, likewise excludes Lighthaven.
The mapped inundation from the modeled northwest and southwest sunny-day breaches follows corridors south of the campus toward the Claremont and Temescal neighborhoods [EBMUD 2021](https://www.ebmud.com/application/files/8316/5729/9817/Claremont_Dam_Mapbook.pdf#page=1 "Claremont Dam inundation mapbook, June 2021").
This conclusion concerns the published scenarios; it does not estimate the probability of dam failure or certify every possible water-infrastructure failure.
[Groundwater is separate.]{.marginnote}
Absence of direct coastal inundation does not establish absence of groundwater near a building.
Rainwater can accumulate above a relatively impermeable layer and emerge on sloping ground; USGS describes this mechanism explicitly [USGS 2019](https://www.usgs.gov/water-science-school/science/water-qa-how-can-water-be-flowing-out-a-hill "Water Q&A: How can water be flowing out of a hill?").
Rain-fed shallow groundwater or seepage is therefore a *plausible* local mechanism, but the reviewed evidence does not establish that it caused the Fay House incident or currently affects a particular Lighthaven room.
A site investigation would distinguish groundwater seepage from surface runoff, roof leakage, and plumbing failures.
**Flooding at the property.**
In **December 2005**, a Rose Garden Inn guest described a flooded entrance and a musty Fay building.
The innkeeper, identified as Kevin A, responded publicly by acknowledging “the flooding problem in the Fay House”.
He attributed the problem to heavy holiday-season rain, said the carpet had been extracted daily, and stated that a contractor would investigate after New Year [Rose Garden Inn 2005](https://www.tripadvisor.com/Hotel_Review-g32066-d80956-Reviews-or50-Rose_Garden_Inn-Berkeley_California.html "December 2005 guest review and the innkeeper’s response concerning Fay House flooding").
The manager's acknowledgment makes this more informative than an uncorroborated guest complaint.
It establishes rain-related flooding at one of the property's buildings, while leaving the entry route, water depth, extent of damage, and eventual repair outcome unresolved.
In an interview published **2025-10-09**, Lighthaven operator Oliver Habryka discussed weather and maintenance at the property, observing: “You will get some small amount of flooding.”
This was a general description of maintenance demands, not a dated incident register or a measurement of flood frequency [Habryka 2025](https://www.complexsystemspodcast.com/episodes/bits-and-bricks-oliver-habryka/ "Bits and bricks: Oliver Habryka on LessWrong, Lighthaven, and community infrastructure").
It gives no basis for assuming that the 2005 entry mechanism persists unchanged after later repairs and renovations.
The public evidence reviewed does not document deep inundation of the entire campus by a creek or the Bay.
It does document a rain-related building incident.
Those are compatible findings: a property can sit outside the major mapped floodplains and still have a consequential water-entry problem.
**Flooding in the neighborhood.**
Berkeley's **2019 Local Hazard Mitigation Plan**, printed page B-126, names **Telegraph Avenue at Stuart Street** among locations where localized flooding can be expected.
The plan says many of the listed Potter watershed locations were confirmed by Public Works staff as chronic nuisance-flooding sites, consistent with experience during the **2004-02-25** storm and the **2005–06** rainy season [Berkeley 2019a](https://berkeleyca.gov/sites/default/files/documents/2019-12-10%20Item%2028%202019%20Local%20Hazard%20Mitigation%20Plan.pdf#page=161 "2019 Local Hazard Mitigation Plan, B-126: Potter watershed flooding locations").
This identifies a municipal drainage concern beside Lighthaven; the wording does not establish that every listed location flooded in every cited storm, or that street flooding entered campus buildings.
Berkeley's longer flood history demonstrates the importance of culvert capacity and blockage.
During **1940-02-27–28**, a tree or stump blocked the Strawberry Creek culvert beneath a pharmacy at Hearst and Euclid, diverting water through the building and into nearby streets.
During **1962-10-12–14**, intense rain and culvert overflow affected Strawberry Canyon, the university campus, International House, and downtown locations [Lawler 1983](https://creeks.berkeley.edu/sites/default/files/publications/lawlerr_1983.historicfloodinginberkeleyespm196.pdf#page=6 "Historic Flooding in Berkeley, pp. 8–11").
These are regional precedents for drainage failure, not records of Lighthaven inundation.
The current hazard plan also records flooding during the **December 2022–January 2023** storms and maps storm-drainage calls from January through June 2023.
Its citywide call map does not establish a specific incident at Lighthaven [Berkeley 2024](https://berkeleyca.gov/sites/default/files/documents/Final%202024%20LHMP%20SENT%20TO%20FEMA%20NOVEMBER%2027.pdf#page=203 "2024 Local Hazard Mitigation Plan, B-195: recent floods; Map 35 at B-187").
Regional histories and citywide service-call maps are useful context, but they cannot replace the property's own incident and repair records.
**Creeks and storm drainage.**
Lighthaven lies within the Potter/Derby drainage system.
The Alameda County Flood Control and Water Conservation District describes a watershed of approximately **9.8 km² (3.8 mi²)**, with almost all of Potter and Derby creeks replaced by constructed underground drainage channels and outlets near Aquatic Park [ACFCD n.d.](https://acfloodcontrol.org/resources/explore-watersheds/potter-derby-creeks-and-powell-street-watersheds/ "Potter/Derby Creeks and Powell Street Watersheds").
UC Berkeley's historical account places Derby Creek's sources near the head of Dwight Way and the Clark Kerr campus, joining near College and Derby and following the general Derby Street corridor toward the Bay [UC Berkeley n.d.](https://creeks.berkeley.edu/node/240 "Derby Creek: History").
That corridor lies north of Lighthaven.
The account is a historical description, not an as-built drawing of today's pipes.
Berkeley's historical creek mapping does not show a creek crossing the campus, but that is a narrower finding than the absence of every buried drainage structure.
Statements that neighboring development sites are not near a mapped creek cannot certify Lighthaven's underground conditions or its legal status.
The City's application checklist explicitly recognizes that a protected creek or culvert may not yet have been identified in City records [Berkeley 2023](https://berkeleyca.gov/sites/default/files/documents/Zoning%20Project%20Application%20Submittal%20Requirements_2023%20Update.pdf#page=10 "Zoning Project Application Submittal Requirements: creek-protection documentation").
[Local hydraulic evidence.]{.marginnote}
The City's 2011 study modeled 488 pipes in the Potter watershed; 271 reached or exceeded full capacity in its 10-year design storm.
Figure B-1 marks **Telegraph/Stuart** in the **74–617 m³ (0.06–0.50 acre-foot)** overflow-volume class; the pipe immediately south lies in the **1.01–2.00** flow-to-capacity class.
These are historical model results, not measured flood depths at Lighthaven [Balance Hydrologics 2011](https://berkeleyca.gov/sites/default/files/2022-03/App.D_DraftPotter_CodornicesWatershedsHydrology_HydraulicsReport_DRAFT_2011July26.pdf#page=39 "Draft Potter and Codornices Watersheds Hydrology and Hydraulics Report, Figure B-1; results on printed pp. 19–20").
The study retains overflow as ponding at each modeled node and returns it there; it does not simulate alternative overland routes along streets.
Its conservative synthetic storm and estimated missing pipe data further limit parcel-level interpretation.
Table A-1's Stuart/Telegraph catchment is **15.1 ha (37.3 acres)**, identifier `36_C`; `MAP22_10YR` identifies rainfall input.
The table's **0.02 is impervious-surface roughness, not a 2% slope** [Balance Hydrologics 2011](https://berkeleyca.gov/sites/default/files/2022-03/App.D_DraftPotter_CodornicesWatershedsHydrology_HydraulicsReport_DRAFT_2011July26.pdf#page=21 "Model limitations on printed p. 17; catchment inputs in Table A-1, PDF p. 37").
The model supports a nearby bottleneck; it cannot show that overflow safely bypasses campus buildings or establish their annual damage probability.
**Infrastructure proposals and status.**
The 2011 Watershed Management Plan proposed **two green-infrastructure/storage units at Telegraph/Stuart**, with a historical planning estimate of $1,158,000.
The project ranked **44th of 44 entries** in the original Potter watershed capital-improvement priority list [Berkeley 2011](https://berkeleyca.gov/sites/default/files/2022-03/WatershedMgtPlan_2011October_Version1.0.pdf#page=89 "2011 Watershed Management Plan, Table 8.1: Potter watershed capital-improvement priorities").
That estimate is not evidence of a current budget, funding commitment, or construction contract.
The adopted **2019 Green Infrastructure Plan** retained Telegraph/Stuart among potential projects and assigned it a **13.5/15 multiple-benefit score**.
The score incorporates benefits and project attributes beyond flood reduction; it is not a flood-probability or damage-severity rating for Lighthaven.
The plan selected **Woolsey Street at Tremont Street**, a different site, as the first proposed storage project for construction [Berkeley 2019b](https://berkeleyca.gov/sites/default/files/2022-04/Green%20Infrastructure%20Plan_Part1.pdf#page=26 "2019 Green Infrastructure Plan, printed pp. 15–17: potential storage projects and prioritization").
The reviewed records do not verify construction or funding of the **Telegraph/Stuart** proposal.
It should therefore remain a historical proposal in this assessment, not installed protection credited to the campus.
As of the assessment date, Berkeley's new **Stormwater Master Plan** remained in planning, collecting drainage-hotspot information and preparing updated infrastructure priorities [Berkeley n.d.](https://berkeleyca.gov/your-government/our-work/capital-projects/stormwater-master-plan "Stormwater Master Plan: current project status").
Berkeley's broader drainage network also has age and capacity constraints: the City's hazard summary describes approximately **150 km (93 mi)** of public storm-drain pipelines, with many assets nearing or beyond their expected useful life and most drainage designed around a 10-year storm [Berkeley 2025a](https://berkeleyca.gov/sites/default/files/2026-06/Berkeley%20Hazard%20of%20Concern-%20Accessible%20Text%20Version.pdf#page=16 "Berkeley’s Hazards: stormwater infrastructure and floodplains").
These citywide conditions support checking the local network; they do not establish the condition of an individual campus connection.
**Present-day interpretation.**
Several mechanisms could produce water damage at this site, and they require different remedies.
A clear roof drain does not resolve an overloaded public pipe; a working sump does not establish that surface runoff cannot cross a doorway.
The following are mechanisms to distinguish, not diagnoses of current defects.
| Mechanism | Site-specific question | Evidence that would resolve it |
|:--|:--|:--|
| Street runoff | Can water escaping Telegraph or Stuart drainage reach a campus opening? | Current surface-flow mapping, curb and threshold survey, storm observations |
| Courtyard ponding | Do garden paths, patios, or stairwells collect water faster than their outlets discharge it? | Local grading, inlet/outlet inspection, records during heavy rain |
| Drain surcharge | Can a campus connection back up when the receiving network is full? | Drainage as-built drawings, connection elevations, updated hydraulic results |
| Rain-fed groundwater | Does subsurface water reach lower floors or foundations? | Wet-season observations, seepage investigation, groundwater measurements if warranted |
| Pump dependence | Does any occupied or service space rely on a sump, and what happens if it stops? | Pump layout, discharge route, service records, alarm and power arrangements |
| Building-envelope or plumbing leakage | Is water entering through roofs, walls, or internal services instead of external flooding? | Incident descriptions and targeted building inspection |
Table: A present-day investigation should identify the water-entry mechanism before selecting a remedy; these questions do not assert that each mechanism is active.
The terrain's westward fall gives water a broad downhill direction but does not guarantee a safe route through a developed campus.
A wall, raised path, blocked outlet, or low threshold can control the local outcome.
Conversely, evidence of an old water-entry problem does not establish that it remains after repairs.
The uncertainty concerns the current arrangement of these features, not whether the property is high above the Bay.
The evidence does not justify an annual percentage chance of indoor flooding, a predicted flood depth, a loss estimate, or a ranking as one of Berkeley's safest properties.
A 10-year drainage design standard is not a calculation that a particular room floods once every decade.
The history establishes that rain-related flooding has occurred; the map establishes a favorable classification for the hazards it represents; neither establishes the missing present-day quantity.
**Resolving the remaining uncertainty.**
The highest-value records are the site's own incident history and drainage alterations.
For the 2005 Fay House incident and later events, the useful facts are the affected room or entrance, water source, storm date, water depth, duration, repair, and recurrence after repair.
Photographs and work orders can distinguish a street-water entry from a roof leak far more directly than another regional flood map.
The next step is to relate **finished floors and vulnerable openings** to the surrounding surfaces and drainage outlets.
A survey should identify the lowest entrances, below-grade spaces, courtyard depressions, retaining-wall openings, and routes to discharge.
Where pumping is involved, its dependence on power, maintenance, and an available discharge route belongs in the same assessment.
For the municipal system, the specific unresolved items are subsequent work at Telegraph/Stuart, the status of the proposed storage units, and updated model results for the relevant inlets and connecting pipes.
The required output is the water level and surface route near the property, rather than merely a pipe-capacity percentage.
These records would determine whether the historical concern has been corrected, remains a manageable localized problem, or warrants further engineering work.