Contents

  1. Stormwater Model
  2. Financial Comparison (Corrected)
  3. MCDA Results
  4. Monte Carlo Sensitivity
  5. Adversarial Test
  6. Environmental Models
  7. Data Sources

1Stormwater Model

SCS Curve Number runoff for 7 development scenarios across 5 design storms. The model uses SSURGO-calibrated soil data with 17 soil map units (30 components). Dominant Hydrologic Soil Group: B (64.0% of site area).

Methodology

SSURGO-calibrated SCS-CN method (TR-55). Curve numbers computed from land cover mix per scenario. Dual HSG soils (B/D, A/D) use undrained (worse) group for conservative estimates. Forest CN from TR-55 Table 2-2c (woods, good condition). Open space CN from TR-55 Table 2-2a (good condition, >75% grass). Impervious CN = 98 (paved surfaces, roofs). Design storms from NOAA Atlas 14 for Asheville, NC region. Helene-scale based on observed 13.98 inches from Hurricane Helene (Sept 2024).

Scenario Summary

Scenario Description CN % Cleared % Impervious
AStadium + Housing85.385%65%
A2Stadium Only76.756%44%
CHeavy Housing81.678%55%
C2Medium Housing72.751%33%
DLight Housing + Park66.022%18%
B/JFull Preservation / Research Forest58.70%0%

Runoff by Scenario and Design Storm (gallons)

Scenario 2-yr 24hr
(3.5")
10-yr 24hr
(5.2")
25-yr 24hr
(6.1")
100-yr 24hr
(7.0")
Helene-scale
(13.98")
A Stadium + Housing 2,493,677 4,378,757 5,412,120 6,459,744 14,792,053
A2 Stadium Only 1,723,874 3,377,523 4,321,346 5,294,931 13,315,466
C Heavy Housing 2,141,774 3,934,748 4,933,486 5,952,875 14,169,942
C2 Medium Housing 1,419,714 2,948,062 3,840,185 4,769,664 12,590,966
D Light Housing + Park 977,943 2,279,224 3,072,642 3,915,640 11,320,403
B Full Preservation 586,299 1,623,332 2,294,435 3,026,635 9,850,775
J Research Forest 586,299 1,623,332 2,294,435 3,026,635 9,850,775

Key Finding

Scenario A (Stadium + Housing) produces 14.8 million gallons of runoff in a Helene-scale event — 4.9 million gallons more than the baseline forest (B/J at 9.9M gallons). That is a 50% increase in stormwater volume from clearing 85% of the canopy.

Even the 2-year storm produces 4.3x more runoff under Scenario A than under preservation.

2Financial Comparison (Corrected)

30-year NPV analysis separated into three perspectives: Public (taxpayers), University (UNCA), and Developer. The original -$132M figure conflated all three perspectives. This corrected model isolates who pays and who benefits.

Critical Correction

The original -$132M was the developer's cost, not the public cost. The actual public cost is the $29M subsidy + ~$15M infrastructure = ~$44M. The developer bears the $204M construction cost as a business risk. These are fundamentally different financial exposures.

Three-Perspective NPV at 5% Discount Rate

Scenario Public NPV @5% University NPV @5% Developer NPV @5%
A Stadium (AECOM) -$46.0M +$18.6M -$149.5M
E-H Housing Swap +$43.3M +$15.5M +$14.7M
H1 Hybrid +$32.3M +$21.4M +$14.7M
B No Action +$10.6M $0 $0
J Research Forest +$11.3M +$24.2M $0

Interpretation

Public perspective: The stadium (A) is the only scenario with negative public NPV. Every alternative generates $10M–$43M in positive public value over 30 years.

University perspective: Research Forest (J) actually generates the highest university NPV (+$24.2M) through grants, field station fees, and endowment. The stadium generates +$18.6M but carries extreme risk.

Developer perspective: The stadium is a -$149.5M loss for the developer. Housing swap scenarios generate +$14.7M — the only profitable development path.

3MCDA Results

Multi-Criteria Decision Analysis across 21 objectives with equal weighting. Course of Action (COA) analysis with Pareto dominance testing and robustness assessment.

12-Scenario Ranking (Equal Weights)

Rank Scenario Description Score Pareto Recommendation
1 H Forest Preserved; No Stadium; Housing MC + City 92.4 Dominant PROCEED
2 E Forest as Park; Stadium 53 Birch; Housing MC 84.0 Dominant PROCEED
3 F Forest Preserved; Stadium Brevard Rd; Housing MC 84.0 Dominant PROCEED
4 I No Stadium; Forest Preserved; Non-RE Revenue 83.7 Dominant PROCEED
5 G Forest Preserved; Stadium South Slope; Housing MC 83.2 Dominated CONSIDER
6 J Research Forest (Living Laboratory) 80.7 Dominant CONSIDER
7 B Full Forest Preservation 72.0 Dominated CONSIDER
8 D Light Housing + Community Park 61.0 Dominated ELIMINATE
9 C2 Medium Housing + Buffers 43.4 Dominant ELIMINATE
10 C Heavy Housing on Forest 35.4 Dominant ELIMINATE
11 A2 Stadium Only on Forest 21.2 Dominated ELIMINATE
12 A Stadium + Housing on Forest (AECOM) 17.0 Dominated ELIMINATE

Recommendation Criteria

PROCEED: Robust (top-3 under ALL weight schemes) or top-5 and non-dominated.

CONSIDER: Partially dominated or mid-ranking; merits further analysis.

ELIMINATE: Dominated by 4+ scenarios or consistently bottom-ranked.

Pareto front: C, C2, E, F, H, I, J are all non-dominated on at least one objective combination. However, Scenario A is dominated by 9 of 11 other scenarios — every other option outperforms it on multiple objectives simultaneously.

4Monte Carlo Sensitivity

1,000 random weight draws (uniform distribution across 20 objectives) to test whether the ranking is sensitive to how you weight the criteria. It is not.

Win Frequency (1,000 Draws)

Scenario Wins Win Rate Interpretation
H Forest Preserved; Housing MC + City 1,000 100.0% Wins under every possible weight combination tested
All other scenarios (A, A2, B, C, C2, D, E, F, G, I, J)00.0%Never ranked first under any weight draw

Key Finding: Scenario A Cannot Enter Top 6

Even at 95% financial weight (treating the decision as almost entirely about money), Scenario A cannot enter the top 6. Its structural disadvantages — zero tax revenue (tax-exempt), $29M public subsidy, irreversible forest clearing — create a deficit that no weight combination can overcome.

Minimax Regret

Minimax regret measures worst-case disappointment: "What is the maximum I could regret choosing this scenario?"

Scenario Max Regret (pts) Interpretation
H1 Hybrid 16 Lowest regret — safest choice regardless of priorities
H Forest Preserved; Housing MC + City16Tied for safest choice
E17Near-minimal regret
F17Near-minimal regret
I17Near-minimal regret
... intermediate scenarios omitted ...
A Stadium on Forest 100 Maximum possible regret — worst choice under any weighting
A2 Stadium Only 100 Maximum possible regret

The regret analysis is unambiguous: H/H1 have a maximum regret of 16 points (you can never be very wrong choosing them), while A has a regret of 100 points (you could be maximally wrong choosing it).

5Adversarial Test

The adversarial test gives Scenario A the developer's best case: maximum plausible scores on every qualitative objective while keeping model-derived scores (stormwater, financial, reversibility) locked. Does H still win?

Method

52 score adjustments favoring development scenarios. Model-locked objectives (stormwater, tax revenue, public subsidy, reversibility, floodway compliance) cannot be changed because they are derived from physical or financial models, not opinion.

H vs. A Gap by Weight Scheme

Weight Scheme Original Gap Adversarial Gap Reduction H Still #1?
Equal 78.2 51.4 -34% Yes (H #1, A #11)
Environmental 80.4 59.6 -26% Yes (H #1, A #12)
Financial 74.1 47.3 -36% Yes (H #1, A #11)
Equity 78.0 46.3 -41% Yes (H #1, A #11)
Governance 81.9 53.2 -35% Yes (H #1, A #12)
Resilience 79.6 54.5 -32% Yes (H #1, A #12)
Floodway 81.9 57.4 -30% Yes (H #1, A #12)

Conclusion

H wins under all 7 weight schemes in both original and adversarial scoring. The gap narrows from ~78 to ~40–60 points but never closes. Even giving A a score of 100 on every qualitative objective, H still wins under most weight schemes. The structural handicap of zero tax revenue, $29M public subsidy, and 85% irreversible clearing is uncloseable.

To flip the result, you would need to change physical reality (make the stadium taxable) or add entirely new objectives (e.g., "athletic prestige") weighted heavily enough to overcome the structural gap.

6Environmental Models

Habitat fragmentation, urban heat island, and carbon stock models for each development scenario. Data from Hansen Global Forest Change (30m), Asheville block group HVI, and Appalachian forest carbon literature.

Habitat Fragmentation

Scenario Cleared (ac) Forest Retained Core Habitat (ac) Narrowest Corridor Edge:Area Ratio
A Stadium + Housing 34.9 15% 1.87 60 m 0.34
C Heavy Housing 31.0 22% 5.45 30 m 0.31
A2 Stadium Only 22.3 44% 3.11 30 m 0.28
C2 Medium Housing 20.2 49% 11.68 30 m 0.31
D Light Development 7.8 78% 38.46 180 m 0.20
B/J Preservation 0.0 100% 47.34 480 m 0.17

Urban Heat Island Impact

Scenario Temp. Increase Trees Preserved Cooling NPV (30yr) HVI Change
A Stadium + Housing +1.98°C (+3.57°F) 675 $92,612 +137%
C Heavy Housing +1.82°C (+3.27°F) 989 $135,831 +126%
A2 Stadium Only +1.31°C (+2.35°F) 1,979 $271,662 +90%
C2 Medium Housing +1.19°C (+2.14°F) 2,205 $302,533 +82%
D Light Development +0.51°C (+0.92°F) 3,510 $481,583 +35%
B/J Preservation 0.0°C 4,500 $617,414 -0.3%

Carbon Stock & Sequestration

Scenario CO2 Released (Mg) Social Cost of Release CO2 Preserved (Mg) Yrs to Recapture
A Stadium + Housing 8,673 $442,347 1,913 43.7
C Heavy Housing 7,959 $405,918 2,806 43.7
A2 Stadium Only 5,714 $291,429 5,612 43.7
C2 Medium Housing 5,204 $265,408 6,250 43.7
D Light Development 2,245 $114,490 9,949 43.7
B/J Preservation 0 $0 12,755 0

Carbon Note

Total forest carbon stock: 3,478 MgC (12,755 Mg CO2 equivalent). Scenario A releases 8,673 Mg CO2 with a social cost of $442,347 (at EPA IWG 2024 rate of $51/ton). It would take 43.7 years of regrowth to recapture the released carbon. Annual sequestration preserved under full forest: 27.3 MgC/yr ($5,109/yr value, $100,133 NPV over 30 years).

7Data Sources

All models use publicly available data. No proprietary data was used in any analysis.

Source Description Use
USDA SSURGO Soil Survey Geographic Database — 17 soil map units, 30 components with hydrologic soil groups Stormwater curve number calibration
NOAA Atlas 14 Precipitation Frequency Data Server — 2yr through 100yr design storms for Asheville, NC Design storm precipitation depths
Hurricane Helene Observations Observed 13.98 inches rainfall during Sept 2024 event Extreme storm scenario
SCS TR-55 Urban Hydrology for Small Watersheds — curve number tables for forest, open space, impervious Runoff computation method
FEMA Flood Maps National Flood Insurance Program maps — SFHA, floodway delineation Flood risk assessment, floodway compliance scoring
Buncombe County Parcels County GIS parcel boundaries with ownership, acreage, tax data (134,436 parcels) Alternative site scoring, public land inventory
Hansen GFC v1.12 Global Forest Change — 30m treecover2000 dataset Forest canopy mapping, fragmentation analysis
Asheville HVI Data Block group Heat Vulnerability Index from urban_heat.geojson Urban heat island modeling, HVI projection
Jenkins et al. 2001 Appalachian oak-hickory carbon stocks (biomass equations) Forest carbon stock estimation
Smith et al. 2006 Forest regrowth rates for Appalachian hardwoods Carbon recapture timeline
Ziter et al. 2019 Urban canopy cooling effects — 0.075°C per 1% canopy Temperature increase modeling
Nowak & Greenfield 2018 Urban tree cooling economics — $7/tree/year cooling benefit Monetary value of cooling services
US EPA IWG 2024 Interagency Working Group Social Cost of Carbon — $51/ton CO2 Monetizing carbon release and sequestration
USFS Forest Valuation Ecosystem services valued at $10K/acre/yr for urban forests 30-year NPV of forest services
Buncombe County Tax Office Multifamily assessment rates — $83K/acre/yr Housing tax revenue projections
AECOM Proposal $204M construction cost, 25,000-seat stadium, P3 structure Scenario A baseline costs
Esri Living Atlas LULC 10m land use/land cover timeseries (2017–2023) Historical canopy change validation