The URL can be used to link to this page

Home My WebLink AboutLiterature Review and Data Gap Analysis July 2026 Committee and Contractor ReviewJuly 6, 2026 1 Hilo Bay Resilience and Watershed Management Plan Task 2 Deliverables-Committee and Contractor Review On May 7, 2026, a Draft Literature Review and Data Gap Analysis report submitted by Lynker Corporation was forwarded to members of the Community Advisory Committee with a request to review and comment by the end of May. Five Community Advisory Committee Members provided input which was shared in a document with Lynker Corporation for consideration. Below is a compilation of committee input and contractor responses received July 6, 2026. July 6, 2026 2 COMMENT #1 – LITERATURE REVIEW One Community Advisory Committee Member spotted a reference for a study that does not exist: • Mead, L. H., & Wiegner, T. N. (2010). Surface water quality along the longitudinal continuum of Canoes-Waiākea Stream, Hilo, Hawaiʻi. Pacific Science, 64(4), 539–549. https://doi.org/10.2984/64.4.539 This member also suggested Lynker conduct a thorough literature search in the UH Hilo Hawaiian collection as well as the TCBES MS repository. CONTRACTOR RESPONSE Regarding misreported reference: We apologize for this oversight. The in-line citation of Mead and Lucas (2010) in reference to the relationship between discharge in Wailuku and turbidity in Hilo Bay was correct. There must have been a copy and paste error, or an error in the reference management software. Either way, it was a lack of careful revision on our part. Regarding recommendation to review UH Hilo Library: A good recommendation. Our timeline, however, has restricted our search to peer- reviewed literature, professional reports, published books, and other professional documents. July 6, 2026 3 COMMENT #2 – REPORT STRUCTURE AND CONTEXT Another Committee Member remarked that the document is very raw with no framework or integration. They recommended that there should be a synthesized analysis along with a description of the workplan (to understand the sequence of tasks and roles of the subcontractors) and a preliminary outline of a water quality management plan to understand the context. They also noted that the document focuses on surface flow— anyone analyzing groundwater flow? Anyone analyzing the sensitivity of the receiving waters in terms of circulation patterns and ecosystem/substrate types. CONTRACTOR RESPONSE Regarding report structure and groundwater focus: It's a good comment to better incorporate the work of other subcontractors into our plan. We've begun to do that by addressing the data gaps. More will come as our specific data collection/modeling plans are completed. We also added to section 9.3. We also plan to analyze the Department of Water Supply’s water quality data and conduct additional sampling of groundwater if necessary. Our sampling map identifies springs (groundwater) to sample in the southern (near the Golf Course), eastern (Keaukaha), and western (Kaumana) portions of the watershed. Circulation patterns in the bay are beyond the scope of this study, but we added a section (9.8) identifying data gaps. July 6, 2026 4 COMMENT #3 – ADDITIONAL RESOURCES FOR LITERATURE REVIEW Former Hilo Bay Watershed Advisory Group Member provided several resources (some are duplicates of resources on the project webpage). CONTRACTOR RESPONSE Regarding additional resources: Thank you. Some of these are very useful, however many of the links are broken. July 6, 2026 5 COMMENT #4 – IMPORTANCE OF UPPER WATERSHEDS/LIT SEARCH A Committee Member said there was inadequate consideration of the landcover scenarios in the upper watersheds. The model in the Lynker part of the study does not include the upper reaches of the watersheds, it almost entirely cuts off all the conservation land. The member went on to say, “However, I’d hope that consideration of scenarios for green infrastructure would include scenarios of protecting upper watershed areas in the conservation land, from approximately the 3,000-8,000’ elevations. That is where the highest rainfall exists and where landcover changes will be most important for retaining water. This includes scenarios for installing fence units in the native forest to prevent damage from hooved animals. The status of native vs alien landcover is a critical factor in stormwater scenarios, both from a flooding and sedimentation perspective. There have been studies that show that fenced forests are 25% faster at infiltrating rainwater than forests that are unfenced with feral pigs. Further, studies show that these native forests are 15 times faster at infiltration than bare ground. Thus, including in the model scenarios where the upper elevation forest is protected (where we install fences that protect thousands of acres of native forest) vs scenarios where these forests are left unprotected and converted to non-native forest will be important for quantifying the importance of the green infrastructure options.” CONTRACTOR RESPONSE Regarding the importance of LULC, green infrastructure, infiltration, sediment runoff, and invasive species: Thank you for these comments. You were right to point out this important and missing component of our draft review. Thank you for the many helpful references. We've updated the LULC section to better characterize the role of forests, state that Hilo's Forests provide ecosystem services, and to discuss the threats to water and soil regulation from invasive species. July 6, 2026 6 COMMENT #5 – COMPREHENSIVE TECHNICAL REVIEW CONTRACTOR RESPONSE (see full comment from committee member below) Regarding “The ASU section is the most technically detailed and scientifically rigorous component of the Task 2 deliverable. Its state-of-knowledge synthesis is thorough, the problem framing is grounded in the published literature across multiple decades, and the proposed monitoring network design is well-calibrated to the scale and complexity of the research questions. The section opens with a culturally grounded introduction that acknowledges the Kānaka Maoli (Native Hawaiian) heritage of the watershed — its moku-o-loko and ahupuaʻa divisions, traditional loko iʻa fishpond management, and the historical meanings embedded in stream names like Wailuku (waters of destruction) and Wailoa (broad river). This framing reflects the community-based orientation of the overall planning process and is likely to strengthen community ownership of monitoring outcomes.” thanks :) Regarding SSC monitoring network design and sampling frequency (“recommendations on autosampler configurations for storm sampling to capture event-driven sediment load”): We agree and we are configuring our autosamplers to 'event-based' sampling following Lewis and Eads, 2009. See this information now reflected in the text. For nutrients: However, the HAR water quality regulations (§11-54-5.2) are (regrettably...) concentration not load based, e.g. "geometric mean not to exceed X value, not to exceed X value 10% of time, not to exceed 2% of time", with different values for wet and dry season. Additionally, DOH has expressed a preference for grab samples (at least for nutrient data) over data from samples from autosamplers. So, I think we are taking this recommendation (i.e. doing our best to capture storms - not sure on exact flow/turbidity thresholds) but we do need to target baseflow data as well Regarding OSDS spatial attribution and loading model (“no mechanistic groundwater loading model linking cesspool density to nitrogen and pathogen concentrations at specific monitoring points ... recommends developing a spatially explicit OSDS groundwater loading model using the USGS MODFLOW framework or a simplified nitrogen loading tool. Input data should include cesspool location, soil hydraulic conductivity (NRCS WSS), depth to groundwater (Hawaii DWS well records), and nitrogen effluent loading rates by system type. Validate model outputs against measured spring and nearshore nitrate concentrations.”): Yes, we will do a spatially explicit model to link OSDS to nutrients - original proposal was to use InVEST but worth revisiting what the other potential options are to get at sub- watershed results in any realistic way (beyond/in addition to statistical models). Re: MODFLOW, Mezzacapo et al. 2022 points out multiple deficiencies of MODFLOW relevant to OSDS pollution. There are updated models that can handle salinity July 6, 2026 7 differences, but for any model there are non-trivial issues related to lack of data characterizing spatial heterogeneity of the subsurface to deal with so many assumptions have to be made. Therefore, we will do our best to link empirical monitoring results to spatially explicit drivers. See additions to section 9.8. Regarding continuous turbidity sensor network (“data-sharing and integration agreement with PACIOOS buoy data”): A great suggestion. Thank you. We added a paragraph and figure 6 linking flow in the Wailuku to sediment delivery and turbidity at the PACIOOS station. Regarding fecal indicator bacteria source tracking (“FIBS and microbial source tracking qPCR”): We do not plan to duplicate efforts of existing ongoing monitoring and only budgeted to analyze samples for nutrients and TSS. FIBs and/or qPCR would require additional resources and would be very resource intensive to do broadly. However, we recognize the importance of source attribution in order to develop realistic and effective intervention strategies. We expect the nutrient and sediment data from our monitoring network will provide a solid foundation to identify locations that can be targeted with specific follow-up studies where source attribution with specific tracers would be most beneficial, and we will seek additional funding to support such analyses. Regarding groundwater discharge quantification (“compute bay-wide groundwater nitrogen and phosphorus loading rates”): J. Oshun has a paper in review quantifying groundwater leakage from upland watersheds to the basal aquifer. This approach might be used with point measurements to scale up nutrient loading. Additional analyses (radon, radium) would be a great addition. We would be interested in collaborating with scientists with access to that methodology/equipment. Regarding nutrient load estimation via storm event sampling (“... storm event nutrient load estimates”): We are doing this but need to be strategic to keep ourselves within budget for the nutrient analyses. We agree the event driven flux needs to be quantified -- with uncertainty -- for the Plan Regarding adaptive monitoring framework and data management (... data management and sharing ”): We've designed data collection SOPs based on recommendations and guidelines from DOH. Because of the limited duration of the project (and based on discussion with DOH), we are not doing a full formal QAPP but are/will maintain clear documentation of lab- specific procedures and align as closely as possible to DOH's draft watershed monitoring July 6, 2026 8 QAPP that is currently being prepared that has been shared with us. We will share data with the USGS when we conclude the project. If there was a strong desire (i.e. from Beth) to have more 'real-time' data available, it would be relatively easy to make a (SIMPLE) github pages RShiny dashboard with a leaflet map and do periodic updates (Hawaii Wai Ola has something similar on their website). Regarding community and indigenous knowledge integration (“community-based monitoring component”): We agree that our data collection could lead directly into a permanent monitoring plan involving the community and other partners (Hawaii Wai Ola). We'll work to continue including Native Hawaiian place names. Regarding cost-benefit analysis framework for BMP alternatives: We'll consider this once our results point to clear interventions. Regarding EPA Section 319 documentation and reporting (“align with EPA 9 required elements”): Yes, a great suggestion. We aim to address these 9 points and have members on our team with experience successfully crafting watershed management plans. Regarding Implementation roadmap (“Recommended actions are grouped into four implementation windows: Immediate (prior to Task 5b kickoff), Short-Term (during Task 5b model construction, 0–12 months), Medium-Term (Year 1–2 monitoring phase), and Long-Term (Year 2–3, watershed management plan integration).” Great suggestions. We will do our best to incorporate what we can into our shorter (~ 1.5 year) time frame. These in-stream data are very difficult to collect, and the data are very valuable, essential, to the watershed management plan. Regarding Implementation roadmap (“The ASU contribution’s integration of Hawaiian cultural context with rigorous hydrologic science is both intellectually compelling and strategically important. By framing the watershed in terms of the moku-o-loko and ahupua’a and acknowledging the deep historical relationship between Kānaka Maoli communities and the bay, ASU has produced a document that speaks to both the scientific community and the broader Hilo community. This framing is likely to strengthen the community-based nature of the overall planning process and to build trust and ownership that will be essential for successful implementation of management recommendations.”) thanks :) Regarding Implementation roadmap (“The geologic-hydrologic analysis in the ASU section — particularly the explanation of how substrate age drives the dramatic differences in surface runoff ratios across the three watersheds — provides a scientifically rigorous foundation for the entire monitoring and modeling program. The conceptual model of the bay as receiving July 6, 2026 9 surface-water-dominated inputs on the west and groundwater-dominated inputs on the east is a powerful and well-supported organizing framework.”) thanks :) COMMUNITY-BASED HILO BAY RESILIENCE AND WATERSHED MANAGEMENT PLAN Comprehensive Technical Review Task 2: Literature Review & Data Gap Analysis May 8, 2026 The goal of this review is not to identify deficiencies in the Task 2 deliverable, which represents a comprehensive and methodologically sound foundation for the overall planning effort, but rather to provide a forward-looking assessment that can strengthen subsequent project phases, particularly Task 5b Hydraulic Modeling and Flood Hazard Maps, the water quality monitoring implementation, and the final watershed management plan. Each thematic section of this document first characterizes what the Task 2 deliverable establishes, then identifies where enhancements or further development would yield the greatest planning benefit. Technical recommendations are drawn from federal agency guidelines, peer-reviewed literature, and established best practices in watershed science, flood hazard management, and community- based environmental planning. Recommendations are assigned a priority level (High, Medium, or Low) reflecting the potential impact on planning outcomes and an effort-to-address estimate to assist the project team in sequencing work within the remaining project budget and schedule. Review Comment Summary Code Section Topic Priority Effort RC-01 EA-Modeling Terrain resolution in urban channels HIGH Moderate RC-02 EA – Modeling Bridge and culvert data collection protocol HIGH Moderate RC-03 EA – Modeling Honoliʻi Stream hydrologic analysis method HIGH Moderate RC-04 EA – Modeling Future climate flow projection methodology HIGH High RC-05 EA – Modeling Model calibration using high- water marks MEDIUM Moderate July 6, 2026 10 Code Section Topic Priority Effort RC-06 EA – Modeling Green infrastructure effectiveness quantification MEDIUM High RC-07 EA – Dat Review Stream bathymetry acquisition approach HIGH Moderate RC-08 EA – Data Review NOAA topo-bathymetric data integration MEDIUM Low RC-09 EA – Data Review Levee and flood control structure condition MEDIUM Moderate RC-10 REDI – Ecological Native vegetation and restoration suitability MEDIUM Moderate RC-11 REDI – Ecological Feral ungulate disturbance and sediment connectivity HIGH Moderate RC-12 REDI – Ecological Coral reef and nearshore habitat linkage MEDIUM Low RC-13 ASU – Water Quality SSC monitoring network design and sampling frequency HIGH Moderate RC-14 ASU – Water Quality OSDS spatial attribution and loading model HIGH Moderate RC-15 ASU – Water Quality Continuous turbidity sensor network integration MEDIUM Low RC-16 ASU – Water Quality Fecal indicator bacteria source tracking MEDIUM Moderate RC-17 ASU – Water Quality Submarine groundwater discharge quantification MEDIUM High RC-18 ASU – Water Quality Nutrient load estimation via storm even sampling HIGH Moderate RC-19 Cross-Cutting Adaptive monitoring framework and data management HIGH High RC-20 Cross-Cutting Community and indigenous knowledge integration HIGH Moderate RC-21 Cross-Cutting Cost-benefit analysis framework for interventions MEDIUM High RC-22 Cross-Cutting EPA Section 319 documentation and reporting alignment HIGH Low Data Gaps and Planned Mitigations EA identifies five key data gaps: bridge geometry, culvert presence and size, stream bathymetry, AEP flow rates for Honoliʻi/Wailuku/Alenaio, and future climate flow rates. The associated mitigation strategies — aerial imagery approximation for structures, bathymetric proxying from LiDAR water surface elevations, drainage area scaling from gauged data, and literature review for climate projections — are professionally reasonable interim approaches. EA appropriately July 6, 2026 11 concludes that despite these gaps, adequate information exists to construct reliable hydraulic models and commits to clearly documenting all assumptions. One methodological concern deserves explicit attention: the memorandum states that FEMA’s effective model flood hazard area maps will serve as the ‘primary reference’ for calibration. This approach introduces a degree of circularity when the updated model is intended to improve upon, and potentially supersede, those same FEMA estimates. The effective models include HEC-1 printouts, HEC-2 files, and HEC-RAS files of varying vintage — some representing methodologies that predate 2D hydraulic analysis by decades. Using these as primary calibration targets risks anchoring new model performance to outdated flood estimates. Primary calibration should instead be grounded in observed flood events, streamgage peak flow records, and documented high-water marks, with FEMA maps serving as a secondary consistency check. This is addressed in detail in RC-05 below. RC-01 Terrain Resolution in Urban Channels and Floodplains—HIGH PRIORITY Observation: The merged LiDAR surveys (2013, 2018, 2023) will serve as the terrain input for HEC-RAS 2D models. However, urban floodplain environments in lower Wailuku, Alenaio, and the Waiākea/Wailoa corridor contain dense infrastructure, subsurface drainage systems, and constructed channels whose hydraulic conveyance is not fully captured by standard LiDAR point clouds. Resolution gaps in LiDAR beneath vegetation canopy and at channel banks can systematically underestimate flood depths in vegetated riparian corridors. The 2013 dataset is now over a decade old and may not reflect post-development changes in the lower watershed. Recommended Enhancement: Conduct a targeted supplemental survey (structure-from- motion photogrammetry or terrestrial LiDAR) of the lower 2–3 km of each modeled stream reach prior to hydraulic model construction. Establish a minimum mesh cell size of 3–5 meters in the active channel and 10–15 meters in the overbank floodplain to resolve urban hydraulic features. Where the 2013 LiDAR is the only coverage, identify areas of significant land use change using Google Earth historical imagery (2013–2025) and flag those areas for manual terrain correction in the HEC-RAS geometry. Apply USACE Engineering Manual EM 1110-2-1416 guidance on terrain processing for 2D models. Supporting References: USACE EM 1110-2-1416 (River Hydraulics); Brunner, G.W. (2016), HEC-RAS 2D Modeling User’s Manual, USACE HEC; McKean et al. (2009) Remote sensing of channel morphology and hydraulic rating curves in mountain rivers, Water Resources Research; FEMA (2020), Guidance for Flood Risk Analysis and Mapping: Hydraulic Models. RC-02 Bridge and Culvert Data Collection Protocol—HIGH PRIORITY July 6, 2026 12 Observation: Limited bridge geometry and uncertain culvert presence are among the most consequential data gaps for model accuracy. Bridges and culverts are primary flow controls in the urban lower watershed and omitting them or approximating them incorrectly can shift 100-year floodplain boundaries by tens to hundreds of meters. The proposed mitigation of approximating geometry from Google Earth and terrain data is a reasonable interim approach but is insufficient for final flood hazard mapping intended to support regulatory and planning decisions. Recommended Enhancement: Develop a structured field data collection protocol for all bridges and culverts within each modeling domain prior to the Task 5b model build. At minimum, record deck thickness, deck elevation (top and bottom chord), span length, pier dimensions, roadway profile, and inlet/outlet condition for culverts. Coordinate with the County of Hawaiʻi Department of Public Works for as-built drawings and inspection records. Apply USACE Engineering Manual EM 1110-2-1603 guidance on hydraulic structure representation in HEC-RAS. For bridges with scour potential, particularly Wailuku River crossings, flag for FHWA HEC-18 scour analysis as a secondary deliverable recommendation for future project phases. Supporting References: USACE EM 1110-2-1605 (1987) (Hydraulic Design of Navigation Dams); Richardson & Davis (2001), Evaluating Scour at Bridges (HEC-18), FHWA; Brunner (2016), HEC-RAS 2D Modeling User’s Manual; FEMA (2016), Bridge Modeling in HEC-RAS Technical Note. RC-03 Honoliʻi Stream Hydrologic Analysis Method—HIGH PRIORITY Observation: Honoliʻi Stream has no FEMA flood hazard mapping, meaning no prior hydrologic study has established AEP peak flows for this watershed. The USGS gage (#16717000) measures discharge in the upper watershed near the forest boundary, not at the channel outlet to Hilo Bay. The coastal plain contributing area below the gage — including urban land cover additions from Kaumana, Puainako, and lower Honoliʻi — is ungaged and potentially significant. Simple drainage area scaling from the gage to the outlet may underestimate peak flows in extreme events due to increasing impervious cover in the lower contributing area. Recommended Enhancement: Develop an explicit hydrologic analysis for Honoliʻi Stream that accounts for the urban contributing area below gage #16717000. Apply both USGS StreamStats regression equations calibrated for east Hawaiʻi Island and an HMS or equivalent rainfall-runoff model using NOAA Atlas 14 precipitation frequency estimates. Use the curve number approach (TR-55/TR-20) to partition pervious and impervious runoff contributions from the lower basin. Cross-validate estimated AEP flows against the drainage area scaling method and document the reconciliation. This is especially important because July 6, 2026 13 Honoliʻi carries documented large sediment loads and fecal indicator bacteria associated with pig activity in the upper watershed. Supporting References: USGS (2016), StreamStats User Guide; NOAA (2011), Atlas 14: Precipitation Frequency Estimates for Hawaii; USDA NRCS (2010), Time of Concentration (TR- 55); Strauch et al. (2014), Climate change and land use drivers of fecal bacteria in tropical Hawaiian rivers, Journal of Environmental Quality; Young & Godszak (2008), Impact of Urbanization on the Quality of Hilo Bay RC-04 Future Climate Flow Projection Methodology—HIGH PRIORITY Observation: The modeling framework includes an ‘Existing under Future Climate Conditions (2075)’ scenario, but EA notes that limited literature addresses projected changes in runoff in this region. Climate projections for the Hawaiian Islands show divergent trends by elevation and season, with higher-elevation rainfall potentially decreasing under trade wind weakening scenarios while low-elevation extreme precipitation events may intensify. Without a robust and explicitly documented method for scaling current AEP flow rates to 2075 conditions, the Future scenario may produce outputs with poorly characterized uncertainty ranges that complicate regulatory and planning use. Recommended Enhancement: Adopt the NOAA/USGS climate-adjusted flood frequency framework for scaling current AEP flows to 2075 conditions. Supplement with the USGS 2011 Water Budget Model rainfall and runoff projections for the Island of Hawaiʻi, which include climate-scenario sensitivity analysis. Coordinate with the University of Hawaiʻi School of Ocean and Earth Science and Technology (SOEST) and the Pacific Islands Climate Adaptation Science Center (PICASC) to obtain island-scale downscaled climate projections consistent with the Hawaiʻi Climate Change Mitigation and Adaptation Commission’s planning guidance. Document the uncertainty range of the 2075 flow projections and propagate that uncertainty into Future scenario flood hazard map outputs using a sensitivity analysis. Supporting References: Hirsch & Ryberg (2012), Has the magnitude of floods across the USA changed with global CO2 levels?, Hydrological Sciences Journal; Hawaiʻi Climate Change Mitigation and Adaptation Commission (2017, updated 2025); Frazier et al. (2016), Comparison of geostatistical approaches to spatially interpolate month-year rainfall for the Hawaiian Islands; USGS (2011), Water-Budget Model and Assessment of Groundwater Recharge for the Island of Hawaiʻi RC-05 Model Calibration Using High Water Marks and Historical Gauge Records—MEDIUM PRIORITY Observation: The memorandum states that FEMA’s effective model flood hazard area maps will serve as the primary calibration reference, supplemented by historical records and July 6, 2026 14 stakeholder observations. Relying primarily on effective FEMA models for calibration is circular when those models are also the baseline for scenario comparison. For Wailuku River, the USGS gage at Piʻihonua (#16704000) has a robust long-term record including multiple large flood events, and the downstream Hilo gage (#16713000, active 1977–1983) provides a historical basis for routing calibration. The November 1–2, 2000 flood event documented as the flood of record for Waiākea provides a specific large-event benchmark. Recommended Enhancement: Develop a formal calibration and validation plan prior to Task 5b model construction that identifies all available high-water mark data, USGS peak flow records at Piʻihonua with routing to the outlet using the established regression relationship (Q_Hilo = 1.48 Q_Piʻihonua + 1.08, R² = 0.98), and the November 2000 peak flow record at Waiākea. Set explicit calibration targets for peak flow, water surface elevation, and flood extent. Apply USACE HEC-HMS/HEC-RAS calibration protocols and report model performance metrics in the Task 5b technical memorandum Supporting References: USACE HEC (2010), HEC-HMS Technical Reference Manual; FEMA (2019), Flood Risk Mapping Guidance — Hydraulic Analyses; Presley et al. (2008), Suspended-sediment and nutrient loads for Waiakea and Alenaio Streams, USGS OFR 2007- 1429 RC-06 Quantifying Green Infrastructure Effectiveness in the Proposed Conditions Scenario—MEDIUM PRIORITY Observation: The Proposed Conditions scenario incorporates green stormwater infrastructure (GSI) adaptations, with model inputs to be provided by other team members. The hydraulic effectiveness of GSI measures in extreme rainfall events — the very events that drive floodplain boundaries — is highly variable and often modest at the watershed scale. Without explicit performance quantification criteria, the Proposed Conditions scenario may overestimate or underestimate GSI benefits, creating unrealistic expectations for mitigation outcomes in planning documents. Recommended Enhancement: Establish explicit performance criteria for GSI measures: retention volume per unit area, peak flow reduction percentage for the design storm, and spatial distribution relative to the modeled floodplain. Use EPA’s National Stormwater Calculator or SWMM to pre-compute GSI performance at the subcatchment level before incorporating results into HEC-RAS boundary conditions. Ensure that GSI performance assumptions are conservative for the 100-year and 500-year events, where surface ponding and inlet surcharging may significantly reduce effective retention capacity. Supporting References: EPA (2014), National Stormwater Calculator User’s Guide (EPA/600/R-13/085a); Rossman (2015), Storm Water Management Model (SWMM 5.1), July 6, 2026 15 EPA/600/R-14/413b; Ahiablame et al. (2012), Effectiveness of Low Impact Development Practices, Water, Air, & Soil Pollution RC-07 Stream Bathymetry Acquisition Approach—HIGH PRIORITY Observation: Existing LiDAR datasets do not capture underwater channel geometry below the water surface. This gap is particularly significant for Wailuku River, which carries large sediment loads and has an active channel that experiences dynamic bed changes during flood events. Without accurate bathymetric cross-sections, HEC-RAS 2D conveyance calculations in the active channel will be based on assumed or interpolated geometry, potentially producing systematic errors in flood depth and velocity predictions that cascade through all five AEP scenarios Recommended Enhancement: Conduct wading surveys at a minimum of 10–15 representative cross-sections per stream for Wailuku River and Alenaio Stream prior to Task 5b model construction. At cross-sections where wading is unsafe at baseflow, deploy a SonTek FlowTracker or similar acoustic Doppler current profiler (ADCP) during moderate flow conditions. For Wailuku River near the bay outlet — the highest-priority reach for accurate conveyance — acquire boat-mounted single-beam sonar bathymetry coordinated with USACE or the County’s Marine Division. Integrate measured bathymetric cross-sections into the HEC-RAS 2D terrain as breaklines and inner boundary refinements. Supporting References: USACE EM 1110-2-1416 (River Hydraulics); SonTek (2019), FlowTracker2 Handheld-ADV Technical Manual; Hilldale & Raff (2008), Assessing the ability of airborne LiDAR to map river bathymetry, Earth Surface Processes and Landforms RC-08 NOAA Topobathymetric Data Integration for the Hilo Bay Coastal Zone— MEDIUM PRIORITY Observation: The memorandum references NOAA LiDAR datasets but does not explicitly address the integration of topobathymetric data at the interface between the terrestrial watershed and Hilo Bay. The coastal plain and bay mouth — particularly the Wailuku River outlet and the Wailoa Pond/estuary — are the zones where tidal backwater effects, sea level rise, and storm surge interact with riverine flooding. Without high-quality topobathymetric data at this interface, downstream boundary conditions in the HEC-RAS 2D model may not accurately represent the transition from fluvial to coastal flood regimes. Recommended Enhancement: Obtain and integrate the NOAA National Centers for Environmental Information (NCEI) topobathymetric DEM for the Hilo Bay coastal zone (NOAA Digital Coast, Coastal National Elevation Database — CoNED). This dataset merges LiDAR topography with bathymetry to produce a seamless elevation model at the land-sea interface. Use the topobathymetric DEM to define the HEC-RAS 2D model downstream July 6, 2026 16 extent and ensure that tidal boundary conditions propagate realistically into the lower river reaches. For sea level rise scenarios, verify that the bay bathymetry accurately represents shoaling zones and breakwater geometry that affect tide-wave propagation. Supporting References: NOAA (2018), Coastal National Elevation Database (CoNED) Application Layers (NCEI); Barnard et al. (2019), Dynamic flood modeling essential to assess the coastal impacts of climate change, Scientific Reports; FEMA (2015), Guidance for Flood Risk Analysis and Mapping — Coastal Analysis and Mapping. RC-09 Levee and Flood Control Structure Condition Assessment—MEDIUM PRIORITY Observation: The memorandum identifies several USACE levee and floodwall documents for hydraulic model integration. However, hydraulic model representation of levees requires not only geometric data but also condition assessment. Levees not accredited by FEMA must be treated as non-existent for Special Flood Hazard Area mapping purposes, which would materially change floodplain boundaries in the lower Alenaio and Wailoa corridors. The current data review does not address accreditation status. Recommended Enhancement: Conduct a review of current FEMA National Levee Database (NLD) records for all levee and floodwall segments within the modeling domain. Verify accreditation status for each structure and coordinate with USACE Honolulu District and FEMA Region 9 to confirm whether levees qualify for SFHA credit in the updated flood hazard maps. Incorporate the most recent USACE levee inspection reports to assess structural condition. Where levee accreditation is uncertain, develop both ‘with-levee’ and ‘without-levee’ flood hazard map products to bracket the range of planning outcomes for affected parcels and infrastructure. Supporting References: FEMA (2013), Procedure Memorandum No. 64 — Process for Identifying Provisionally Accredited Levees; USACE (2022), National Levee Database (NLD) Technical Reference; USACE (2016), Alenaio Stream Levee and Floodwall Inspection Reports Roth Ecological Design International — Ecological Data and Green Infrastructure Siting The most significant outstanding data needs for REDI are the modeled catchment boundaries and flood extents, which must be delivered by the EA Engineering hydraulic modeling workstream before REDI’s GSI prioritization can be finalized. This creates a sequencing dependency that should be explicitly built into the project schedule. REDI also needs confirmation of the preferred watershed scale (HUC12 vs. HUC8) and applicable TMDL information from HDOH. July 6, 2026 17 Ecological Data Gaps and Enhancements RC -10 Native Vegetation Mapping and Restoration Suitability Assessment— MEDIUM PRIORITY Observation: The REDI data inventory identifies land cover, impervious surface, and tree canopy from NOAA Digital Coast. However, ecological suitability analysis for Nature-Based Solutions interventions requires more granular native versus invasive species mapping than current land cover products provide. NOAA C-CAP classifies vegetation at a broad level (forest, shrub/scrub, grassland) but does not distinguish intact native forest from invasive- dominated stands such as strawberry guava monocultures, which have fundamentally different hydrologic properties and restoration cost profiles. Recommended Enhancement: Commission or obtain a current-conditions native/invasive vegetation classification for the three primary watersheds using either USGS or DLNR vegetation maps derived from WorldView or Planet satellite imagery with ground-truthing, or the Hawaiʻi Statewide Assessment of Forest Conditions (DOFAW, 2016) as the planning baseline, supplemented with targeted field surveys in upper watershed restoration priority areas. Overlay vegetation type with slope, soil type, stream proximity, and ownership to develop a restoration suitability scoring matrix that directly supports prioritization of riparian buffer, reforestation, and bioretention interventions. Supporting References: DOFAW (2016), Hawaiʻi Statewide Assessment of Forest Conditions and Resource Strategy; Cole & Litton (2014), Vegetation response to removal of non-native feral pigs from Hawaiian tropical montane wet forest, Biological Invasions; Asner et al. (2008), Invasive species detection in Hawaiian rainforests using airborne imaging spectroscopy and LiDAR, Remote Sensing of Environment. RC-11 Feral Ungulate Disturbance and Sediment Connectivity—HIGH PRIORITY Observation: The ASU section notes that the NWQI critical source area methodology does not fully capture sediment production from forests inhabited and modified by feral ungulates. Feral pig and goat activity in the upper Honoliʻi and Wailuku watersheds is well documented as a primary driver of soil disturbance, understory destruction, and streambank erosion in native Hawaiian forests. REDI’s ecological data inventory does not explicitly include ungulate distribution or impact mapping, and current NOAA land cover products cannot resolve this Recommended Enhancement: Integrate feral ungulate distribution and impact data from DLNR Division of Forestry and Wildlife pig and goat management unit records, USDA APHIS Wildlife Services feral pig removal data, and field transect data from published studies in the upper watershed. Develop a GIS layer of ‘ungulate disturbance zones’ and overlay with EA Engineering critical source area mapping and ASU’s sediment monitoring network design to July 6, 2026 18 identify reach segments where elevated SSC is likely attributable to ungulate-driven soil disturbance. This distinction is critical for targeting cost-effective interventions: ungulate exclusion fencing in upper watershed headwaters may deliver greater sediment load reductions per dollar invested than downstream urban BMPs. Supporting References: Strauch et al. (2014), Climate change and land use drivers of fecal bacteria in tropical Hawaiian rivers, Journal of Environmental Quality; Cole & Litton (2014), Biological Invasions; DLNR DOFAW (2023 & 2024), Feral Ungulate Management Plan, Wailuku River State Park and Upper Watershed. RC-12 Coral Reef and Nearshore Habitat Linkage to Watershed Loads—MEDIJM PRIORITY Observation: The REDI data inventory covers terrestrial ecological data but does not address the nearshore coral reef and fishpond (loko iʻa) ecosystems of Hilo Bay, which are the ultimate ecological receptors of the sediment, nutrient, and bacterial loads delivered from the watershed. Hilo Bay’s eastern shoreline (Keaukaha) supports coral communities documented as sensitive to turbidity exceedances, and the Wailoa Pond estuary functions as a critical nursery habitat. The watershed management plan’s ecological success metrics should include nearshore habitat condition, not only terrestrial vegetation. Recommended Enhancement: Establish a nearshore ecological baseline monitoring component coordinated with the Division of Aquatic Resources (DAR) and the NOAA Pacific Islands Fisheries Science Center. Obtain existing coral reef survey data from the NOAA National Coral Reef Monitoring Program (NCRMP) for the Hilo Bay sector. Cross-reference spatial distribution of turbidity exceedance events with coral colony locations and benthic cover data. Include coral reef and fishpond habitat condition as co-benefit metrics in the NBS project evaluation framework, enabling quantification of ecological return on investment for riparian and sediment reduction projects. Supporting References: Lucas et al. (2023), Spatially distributed water quality responses to freshwater discharge in a tropical estuary, Hilo Bay, Environmental Monitoring and Assessment; NOAA NCRMP (2021), Pacific Islands Network Coral Reef Monitoring Program; Wiegner et al. (2013), Water quality between low- and high-flow conditions in a tropical estuary, Estuaries and Coasts. Arizona State University — Watershed Hydrology and Water Quality State of Knowledge The ASU section is the most technically detailed and scientifically rigorous component of the Task 2 deliverable. Its state-of-knowledge synthesis is thorough, the problem framing is grounded in July 6, 2026 19 the published literature across multiple decades, and the proposed monitoring network design is well-calibrated to the scale and complexity of the research questions. The section opens with a culturally grounded introduction that acknowledges the Kānaka Maoli (Native Hawaiian) heritage of the watershed — its moku-o-loko and ahupuaʻa divisions, traditional loko iʻa fishpond management, and the historical meanings embedded in stream names like Wailuku (waters of destruction) and Wailoa (broad river). This framing reflects the community-based orientation of the overall planning process and is likely to strengthen community ownership of monitoring outcomes. ASU Data Gaps and Monitoring Program ASU identifies six data gaps that collectively define the primary research agenda for the monitoring phase. The proposed sampling network covering Alenaio, Honoliʻi, Keaukaha, Pukihae, Puueo, Waiakea, and Wailuku streams and drainages is well-distributed and appropriate for the study objectives. The combination of autosamplers, turbidity probes, and grab sampling, coordinated with EPA guidelines for HAR 11-54 compliance reporting, represents a professionally designed monitoring program. The plan to submit data to HDOH for the 2028 Integrated Report (§303(d) and §305(b)) provides clear regulatory accountability. The following review comments address specific enhancements to the ASU monitoring program that would substantially strengthen the scientific and regulatory value of the data collected. RC-13 SSC Monitoring Network Design: Sampling Frequency and Event Coverage—HIGH PRIORITY Observation: The proposed network uses turbidity probes and autosamplers to develop SSC-discharge rating curves across the three primary watersheds, consistent with USGS protocols. However, the critical scientific and regulatory value of SSC monitoring in Hilo Bay lies in event-scale (storm) sampling, not baseflow characterization. Wailuku River’s storm- driven SSC response is highly non-linear (supply-limited, power law exponent 0.54), meaning that the largest 5–10% of events by discharge deliver a disproportionate fraction of the annual sediment load. Autosampler deployment must be configured to capture these events at high temporal resolution or the rating curve will be systematically biased toward low-SSC baseflow conditions. Recommended Enhancement: Configure autosamplers to trigger at least every 2 hours during discharge events exceeding the 30th percentile exceedance flow (approximately 3.5 m³/s at Wailuku Piʻihonua, the threshold identified for turbidity exceedance in Hilo Bay). For events exceeding the 10th percentile (high flow), increase sampling frequency to every 30– 60 minutes to capture the rising limb, peak, and falling limb of the sediment hysteresis loop. Program turbidity sensors with a 15-minute logging interval during storms. Submit all SSC- July 6, 2026 20 turbidity and SSC-discharge data to the USGS National Water Information System for long- term archiving and public accessibility. Follow USGS Techniques and Methods 3-C4 (Rasmussen et al. 2009) for rating curve development and uncertainty quantification. Supporting References: Rasmussen et al. (2009), Guidelines and Standard Procedures for Continuous Water-Quality Monitors, USGS Techniques and Methods 1-D3; Gray & Gartner (2009), Surrogate Technologies for Monitoring Bed-Material Sediment Transport in Rivers, USGS; Edwards & Glysson (1999), Field Methods for Measurement of Fluvial Sediment, USGS TWRI Book 3, Chapter C2. RC-14 Spatial Attribution and Groundwater Loading Model—HIGH PRIORITY Observation: The current analysis relies on cesspool count and distribution data without a mechanistic groundwater loading model linking cesspool density to nitrogen and pathogen concentrations at specific monitoring points. This limits the ability to prioritize cesspool conversion under Act 125 compliance by watershed impact. The estimated 5.6 MGD of untreated effluent from over 10,000 OSDS represents a dominant chronic loading pathway, but its spatial attribution to specific monitoring stations and drinking water well capture zones has not been quantified. Recommended Enhancement: Develop a spatially explicit OSDS groundwater loading model using the USGS MODFLOW framework or a simplified nitrogen loading tool. Input data should include cesspool location, soil hydraulic conductivity (NRCS WSS), depth to groundwater (Hawaii DWS well records), and nitrogen effluent loading rates by system type. Validate model outputs against measured spring and nearshore nitrate concentrations. Use model results to prioritize cesspool conversion: high-hydraulic-conductivity, shallow-water- table areas with high cesspool density that discharge to monitored coastal springs or drinking water well capture zones should receive highest priority for County sewer extension under Act 125. Supporting References: Whittier & El-Kadi (2014), Human health and environmental risk ranking of OSDS, University of Hawaiʻi WRRC; Mezzacapo & Shuler (2022), Characterizing nutrient sources and transport in the Wailoa River and Waiākea Pond, UH Mānoa WRRC; Waiki et al. (2025), Sewage pollution from OSDS in coastal waters of Keaukaha, Journal of Hydrology: Regional Studies; EPA (2002), Onsite Wastewater Treatment Systems Manual. RC-15 Integration of Continuous Turbidity Sensor Network with PACIOOS Buoy Data—MEDIUM PRIORITY Observation: The PACIOOS buoy near the Wailuku River mouth provides continuous turbidity, temperature, salinity, chlorophyll, and dissolved oxygen data at 15-minute intervals — a valuable anchor for the proposed monitoring network. The ASU document July 6, 2026 21 references this data source but does not propose an explicit integration protocol between the in-stream turbidity sensors and the buoy. Without a formal data linkage protocol, the relationship between watershed sediment loads and bay water quality outcomes will remain inferential rather than quantified. Recommended Enhancement: Establish a formal data-sharing and integration agreement with PACIOOS at the outset of the monitoring program. Develop a standardized data fusion workflow that pairs in-stream discharge and turbidity measurements with buoy turbidity records on matched time steps. Use the existing regression relationship (Q_Wailuku vs. NTU_buoy; Mead & Wiegner 2010) as the baseline and refine it with new paired measurements. Produce time-series plots of watershed load delivered versus bay turbidity response for each major storm event, enabling quantification of the attenuation or amplification of sediment signals at the bay interface. This paired dataset will be essential for future TMDL development. Supporting References: PACIOOS (2024), Hilo Bay Water Quality Buoy Data Archive. RC-16 Fecal Indicator Bacteria Source Tracking—HIGH PRIORITY Observation: The ASU review documents high concentrations of Enterococcus, Clostridium perfringes, and antibiotic-resistant Staphylococcus aureus in Hilo Bay, with 40–50% of samples testing positive for human waste markers. While the proposed monitoring plan will characterize fecal indicator bacteria (FIB) concentrations in streams and along the coast, the design does not include a microbial source tracking (MST) component that would identify the relative contributions of human (OSDS/sewer), livestock (cattle, pig), and wildlife sources to FIB loading. Without MST data, it is not possible to prioritize interventions between OSDS conversion, agricultural BMPs, and feral animal management. Recommended Enhancement: Add an MST component using quantitative PCR (qPCR) for human-specific Bacteroides markers (BacHum or HF183) to identify OSDS contributions; Pig- 2-Bac markers to distinguish feral pig contributions (documented by Strauch et al. 2014 for Honoliʻi); and ruminant markers for cattle and goat contributions in agricultural sub- watersheds. Sample at a subset of the proposed monitoring stations during both baseflow and storm events. Analyze samples at a state-certified molecular biology laboratory. MST data will provide source attribution required for HDOH Clean Water Act reporting and will be essential documentation for EPA Section 319 grant applications. Supporting References: Harwood et al. (2014), Microbial source tracking markers for detection of fecal contamination in environmental waters, FEMS Microbiology Reviews; Strauch et al. (2014), Journal of Environmental Quality; Gerken et al. (2021), Environmental surveillance of antibiotic resistant Staphylococcus aureus, Antibiotics; EPA (2005), Microbial Source Tracking Guide Document (EPA/600/R-05/064). July 6, 2026 22 RC-17 Submarine Groundwater Discharge Quantification—MEDIUM PRIORITY Observation: The ASU document estimates submarine groundwater discharge (SGD) to the eastern half of Hilo Bay at 476 MGD based on 1980 data using mass balance methods — a figure that predates the extensive OSDS documentation now available and does not account for 45 years of changes in groundwater recharge patterns or land use. More recent SGD measurements in Keaukaha (Waiki et al. 2025) provide localized flow rates but not bay-wide estimates. SGD is identified as the dominant pathway for nutrient and contaminant delivery to the eastern bay, making its accurate quantification a high priority. Recommended Enhancement: Include a targeted SGD quantification effort using radium isotope tracing (²²⁴Ra, ²²³Ra) or radon (²²²Rn) as natural geochemical tracers. This approach has been successfully applied in other Hawaiian coastal settings and is logistically feasible with a small boat and portable gamma spectrometer. Sampling during low-flow, dry-season conditions (May–October) will maximize the SGD signal relative to river-dominated inputs. Pair SGD flux estimates with nutrient concentration measurements in coastal springs to compute bay-wide groundwater nitrogen and phosphorus loading rates. Supporting References: Dulaiova, H., Camilli, R., Henderson, P.B., and Charette, M.A. (2010). Coupled radon, methane and nitrate sensors for large-scale assessment of groundwater discharge and non-point source pollution to coastal waters. Journal of Environmental Radioactivity; Knee et al. (2010), Using radon and radium isotopes to characterize groundwater inputs, Limnology and Oceanography; Waiki et al. (2025), Journal of Hydrology: Regional Studies. RC-18 Storm Event Nutrient Sampling for Load Estimation—HIGH PRIORITY Observation: The ASU monitoring plan focuses on seasonal nutrient sampling following EPA HAR 11-54 criteria. This regulatory-compliance framing is appropriate for §303(d) reporting. However, it may not capture the disproportionate contribution of storm event fluxes to total annual nutrient loads. Published data from Honoliʻi and Waiākea show that nutrient concentrations respond non-linearly to discharge — some streams dilute at high flows while others show chemostatic or mobilization behavior. Without storm event sampling, annual load estimates will rely on extrapolation from baseflow samples, introducing substantial uncertainty into TMDL calculations. Recommended Enhancement: Implement an event-triggered nutrient sampling protocol using autosamplers deployed in parallel with the turbidity sensors. Program autosamplers to collect samples every 1–2 hours during storm events exceeding a discharge threshold corresponding to the 30th percentile exceedance. Analyze samples for the full HAR 11-54 suite: Total Nitrogen, Nitrate+Nitrite, Total Phosphorus, and Total Suspended Solids. Use event-scale concentration-discharge hysteresis analysis to characterize source flushing July 6, 2026 23 versus dilution dynamics. Compute event-specific and annual nutrient loads using the USGS LOAD ESTimator (LOADEST) software to propagate uncertainty from rating curves into load estimates. Supporting References: Presley et al. (2008), USGS OFR 2007-1429; Wiegner et al. (2013), Estuaries and Coasts; Runkel et al. (2004), Load Estimator (LOADEST): A FORTRAN Program for Estimating Constituent Loads in Streams and Rivers, USGS Techniques and Methods 4-A5. Cross-Cutting Recommendations The three Task 2 sections address related phenomena from distinct disciplinary perspectives, and their combined value will depend substantially on the degree to which hydraulic model outputs, ecological data, and water quality monitoring are woven together into a coherent management framework. At present, the three workstreams are described largely in isolation: EA’s hydraulic model generates flood extents that REDI’s GSI siting analysis requires, but there is no explicit protocol for how or when that handoff occurs. ASU’s turbidity monitoring will inform sediment transport dynamics that should update assumptions in EA’s roughness calibration, but that feedback loop is not articulated. REDI’s restoration priority sites should logically coincide with ASU’s monitoring stations to enable measurement of treatment effectiveness, but this co-location is not proposed. The following four cross-cutting recommendations address the structural integration needs that will determine whether the three workstreams cohere into a watershed management plan of lasting value or remain parallel technical deliverables. RC-19 Adaptive Monitoring Framework and Integrated Data Management—HIGH PRIORITY Observation: The three monitoring workstreams will generate datasets that are most valuable when integrated: hydraulic model outputs inform where to place water quality sensors; turbidity data validates sediment transport assumptions in the hydraulic model; ecological restoration sites should be co-located with monitoring stations to measure treatment effectiveness. Without an explicit adaptive monitoring framework and shared data management system, these cross-workstream synergies will be lost and the planning team will be unable to respond dynamically to emerging data during the monitoring phase. Recommended Enhancement: Establish a Project Data Management Plan (DMP) at the outset of the monitoring phase that defines a common spatial reference system (Hawaiʻi State Plane, NAD83) and data schema for all field measurements; specifies data submission timelines (raw data within 48 hours of collection; QA/QC validated data within 30 days); identifies a shared cloud data repository accessible to all team members; and defines July 6, 2026 24 adaptive management trigger points — specific data thresholds that will prompt protocol modifications during the monitoring period. Apply EPA’s Guidelines for Preparing Quality Assurance Project Plans (QAPPs) as the framework for data quality standards. Schedule quarterly data review meetings across all team components Supporting References: EPA (2002), Guidance for Quality Assurance Project Plans (EPA/240/R-02/009); USGS (2006), Office of Water Quality Technical Memorandum 2006.01; EPA (2008), Handbook for Developing Watershed Plans to Restore and Protect Our Waters (EPA 841-B-08-005). RC-20 Community and Indigenous Knowledge Integration—HIGH PRIORITY Observation: The ASU introduction provides a culturally grounded framing of the watershed, acknowledging Kānaka Maoli place names, traditional resource management practices, and the historical relationship between the Hilo community and the bay. However, the monitoring and modeling tasks are described in exclusively Western scientific terms. Indigenous ecological knowledge (IEK) and community-based participatory monitoring have been demonstrated to strengthen watershed monitoring programs in Pacific Island settings by extending spatial coverage, identifying monitoring priorities, and building long-term community ownership of data and management outcomes. Recommended Enhancement: Formally integrate a Community-Based Monitoring (CBM) component into the ASU water quality monitoring program. Partner with existing community organizations — Hawaiʻi Wai Ola, Wailoa River and Recreation Area stakeholders, and Native Hawaiian practitioner groups managing Wailoa Pond — to co- design a subset of sampling locations and protocols reflecting community water quality priorities such as swimming safety, traditional fishing areas, and drinking water springs. Train community monitors in basic water quality sampling using low-cost sensors and submit results to the shared data platform alongside professional monitoring network data. Acknowledge Indigenous place names and traditional water resource management frameworks in all public-facing documents and the final watershed management plan. Supporting References: EPA (2016), Environmental Justice Collaborative Problem-Solving Cooperative Agreement Program; Silvius (2005), Hilo Bay Watershed Management Plan; Maly & Maly (2003), Panaeʻo: A Cultural-Historical Study of the Hilo Bay Watershed; Berkes (2009), Evolution of co-management: Role of knowledge generation, bridging organizations and social learning, Journal of Environmental Management. RC-21 Cost Benefit Analysis Framework for Watershed Interventions—MEDIUM PRIORITY Observation: ASU Task 2 gap #4 identifies the need for a cost-benefit analysis to prioritize interventions. The watershed management plan will ultimately need to present a prioritized July 6, 2026 25 action plan to the County, the State, and federal funding agencies. Without a consistent economic framework, comparison across intervention types — cesspool conversion, riparian fencing, GSI, levee maintenance — will rely on qualitative judgment rather than quantified return on investment, weakening the plan’s credibility with funding agencies. Recommended Enhancement: Adopt the EPA Handbook for Developing Watershed Plans framework for cost-effectiveness analysis of BMP alternatives. For each proposed intervention, quantify capital cost, annual operation and maintenance cost, pollutant or sediment load reduction in kg per year, cost per unit load reduction in dollars per kg per year, and co-benefits including flood peak reduction, habitat improvement, and community access. Apply USDA NRCS RUSLE2 to estimate sediment reduction from agricultural and forest BMPs. Present results as a cost-effectiveness ranking table suitable for direct incorporation into the NRCS NWQI implementation plan and future FEMA BRIC grant applications. Supporting References: EPA (2008), Handbook for Developing Watershed Plans (EPA 841-B- 08-005); USDA NRCS (2013), RUSLE2 Science Documentation. RC-22 EPA Section 319 Documentation and Reporting Alignment—HIGH PRIORITY Observation: The ASU monitoring plan is explicitly designed to support HDOH Clean Water Act §303(d) reporting for the 2028 Integrated Report deadline. However, the plan does not address the parallel documentation requirements for EPA Section 319 (Nonpoint Source Pollution) grant eligibility, which is likely to be the primary federal funding mechanism for watershed management implementation. Section 319 requires a state-approved watershed- based plan (9 Elements Plan) as a prerequisite for implementation funding, and several of the 9 Elements have direct data requirements that must be addressed during the Task 2 monitoring phase. Recommended Enhancement: Align the Task 2 monitoring and modeling outputs with the 9 Elements of an EPA Watershed-Based Plan as required for Section 319 funding eligibility. The specific data requirements that intersect with Task 2 are: Element 1 (identification of pollutant sources and loads, addressed by ASU monitoring); Element 2 (expected load reduction under management measures, addressed by ASU and EA modeling); Element 3 (management measures to achieve load reduction, addressed by REDI NBS design); Element 5 (technical and financial assistance, requiring a funding matrix); and Element 8 (monitoring to track progress, addressed by the ASU monitoring plan). Begin pre-application coordination with the Hawaiʻi Department of Health Clean Water Branch and USEPA Region 9 Water Division immediately to confirm that Task 2–5 deliverables will satisfy 9 Elements requirements for the 2028 application cycle. July 6, 2026 26 Supporting References: EPA (2008), Handbook for Developing Watershed Plans (EPA 841-B- 08-005); 40 CFR Part 130 (Water Quality Planning and Management); USEPA Region 9 (2023), Section 319 Program Implementation Guidance for Pacific States and Territories; Hawaiʻi Department of Health Clean Water Branch (2024), 2024 Nonpoint Source Management Program Annual Report. Implementation Roadmap The following sequencing integrates all review comments into a phased approach aligned with the project’s existing Task 2–Task 5b workflow. Recommended actions are grouped into four implementation windows: Immediate (prior to Task 5b kickoff), Short-Term (during Task 5b model construction, 0–12 months), Medium-Term (Year 1–2 monitoring phase), and Long-Term (Year 2– 3, watershed management plan integration). Immediate Actions (Prior to Task 5b Kickoff) • RC-02: Initiate bridge and culvert field survey coordination with County of Hawaiʻi Department of Public Works; request as-built drawings and inspection records. • RC-09: Query FEMA National Levee Database and request USACE levee inspection reports for all Alenaio and Wailoa levee and floodwall structures; confirm accreditation status. • RC-22: Schedule pre-application meeting with HDOH Clean Water Branch and EPA Region 9 to align Task 2–5 deliverables with EPA Section 319 Nine Elements documentation requirements for the 2028 application cycle. • RC-19: Draft Project Data Management Plan and Quality Assurance Project Plan (QAPP) for the ASU monitoring program; establish shared cloud data repository and common data schema. • RC-20: Initiate contact with Hawaiʻi Wai Ola and Wailoa River stakeholder groups to co- design Community-Based Monitoring sampling locations and protocols. Short-Term Actions (Task 5b Model Construction, 0–12 Months) • RC-01: Conduct supplemental structure-from-motion photogrammetry or terrestrial LiDAR survey of the lower 2–3 km of Wailuku and Alenaio stream channels to fill terrain resolution gaps. • RC-07: Deploy wading survey team for bathymetric cross-sections at 10–15 priority cross-sections per stream on Wailuku River and Alenaio Stream; coordinate boat- mounted sonar survey at Wailuku River bay outlet. • RC-03: Complete Honoliʻi hydrologic analysis using combined StreamStats, curve number, and drainage area scaling approach; document reconciliation of methods. • RC-05: Develop formal calibration and validation plan; compile Piʻihonua gage record, November 2000 flood benchmark data, and available high-water marks. July 6, 2026 27 • RC-13: Install turbidity probes and autosamplers at initial monitoring stations; configure storm-event triggering thresholds at 30th percentile exceedance flow. • RC-16: Collect baseline microbial source tracking (MST) samples at watershed outlet stations during the first wet season; analyze for human-specific Bacteroides, Pig-2-Bac, and ruminant markers. • RC-11: Obtain DLNR DOFAW ungulate management unit data; develop GIS ungulate disturbance zone layer and overlay with EA critical source area map. Medium-Term Actions (Year 1–2, Monitoring Phase) • RC-04: Complete future climate flow projection analysis in coordination with UH SOEST PICASC; document uncertainty range and propagate into Future scenario flood hazard maps. • RC-14: Build OSDS groundwater loading model using MODFLOW or EPA Nitrogen Loading Model framework; validate against Waiki et al. 2025 and Mezzacapo & Shuler 2022 coastal monitoring data. • RC-17: Deploy radon or radium SGD sampling transects during dry season; compute bay- wide groundwater nutrient loading rates for nitrogen and phosphorus. • RC-18: Begin storm event nutrient sampling with autosamplers; compute event loads using USGS LOADEST; develop annual load estimates. • RC-10: Commission native/invasive vegetation classification for upper watershed priority restoration zones; develop restoration suitability scoring matrix. • RC-21: Develop cost-effectiveness ranking table for proposed watershed interventions using EPA framework and RUSLE2 sediment reduction estimates. Long-Term Actions (Year 2–3, Watershed Management Plan Integration) • RC-06: Incorporate GSI performance data from SWMM pre-modeling into Proposed Conditions HEC-RAS scenario boundary conditions. • RC-08: Obtain CoNED topobathymetric DEM for Hilo Bay coastal zone; refine downstream model boundary conditions for sea level rise scenarios. • RC-12: Obtain NOAA NCRMP coral reef survey data for Hilo Bay; develop nearshore habitat co-benefit metrics for NBS project evaluation framework. • RC-15: Develop formal PACIOOS buoy data integration protocol; produce paired watershed sediment load versus bay turbidity response time-series for each major storm event. • Synthesize all monitoring results, hydraulic model outputs, and ecological assessments into the final 9-Elements Watershed Management Plan in compliance with EPA 841-B- 08-005 and Section 319 documentation requirements. Overall Assessment and Conclusions July 6, 2026 28 Strengths of the Task 2 Deliverable Task 2 represents a technically sound, well-coordinated, and appropriately scoped literature review and data gap analysis. Several aspects of the deliverable are particularly noteworthy: The ASU contribution’s integration of Hawaiian cultural context with rigorous hydrologic science is both intellectually compelling and strategically important. By framing the watershed in terms of the moku-o-loko and ahupuaʻa, and acknowledging the deep historical relationship between Kānaka Maoli communities and the bay, ASU has produced a document that speaks to both the scientific community and the broader Hilo community. This framing is likely to strengthen the community-based nature of the overall planning process and to build the trust and ownership that will be essential for successful implementation of management recommendations. The geologic-hydrologic analysis in the ASU section — particularly the explanation of how substrate age drives the dramatic differences in surface runoff ratios across the three watersheds — provides a scientifically rigorous foundation for the entire monitoring and modeling program. The conceptual model of the bay as receiving surface-water-dominated inputs on the west and groundwater-dominated inputs on the east is a powerful and well- supported organizing framework. EA’s selection of HEC-RAS 2D with a well-defined four-scenario matrix, combined with explicit acknowledgment of data gaps and documentation commitments, reflects professional engineering practice at the appropriate level for this type of watershed-scale planning study. REDI’s data needs memo, while brief, is appropriately structured for a planning-phase deliverable and demonstrates a well-curated baseline dataset that spans the necessary spatial and thematic coverage for a parcel-level GSI siting analysis. Priority Areas for Strengthening The most significant areas where the Task 2 foundation can be strengthened are: The treatment of model calibration deserves special attention. The reliance on FEMA effective models as the primary calibration reference introduces a circularity risk that could compromise the credibility of the updated flood hazard maps, particularly for Honoliʻi Stream where no FEMA mapping currently exists. A calibration strategy grounded in observed peak flows, high- water marks, and the well-documented November 2000 flood event will produce more defensible outputs. The future climate flow projection approach — currently described only as a literature review plus EA internal resources — represents the most analytically consequential gap for the 2075 July 6, 2026 29 scenario modeling. The divergent and potentially counterintuitive direction of climate change effects on Hawaiian Island hydrology (possible decreasing trade wind rainfall at high elevations, intensifying extreme events at low elevations) demands a more structured and explicitly documented projection methodology. The cross-workstream integration challenge is structural rather than technical. The three Task 2 sections are individually well-executed, but the mechanisms by which they will be synthesized — in terms of data exchange, joint interpretation, and integrated recommendations — are not yet specified. The implementation of a Project Data Management Plan, shared data repository, and adaptive monitoring framework will be essential to realizing the full value of the combined program. Finally, the EPA Section 319 alignment requirement should be addressed immediately rather than at the end of the planning process. Pre-application coordination with HDOH and EPA Region 9 will ensure that the data collected and the models built during Tasks 2 through 5 satisfy the Nine Elements documentation requirements that will determine eligibility for the federal implementation funding the County will need to act on the management plan’s recommendations. Conclusion The Hilo Bay Watershed faces genuinely complex and deeply intertwined stressors: decades of cesspool-driven groundwater contamination, episodic storm-driven sediment and bacterial loading, sea level rise compounding coastal flood hazard, and ongoing ecological degradation of a reef and estuarine system of extraordinary cultural and ecological significance. The Task 2 approach correctly frames these challenges with scientific rigor and community grounding. The 22 review comments in this document are offered in the spirit of continuous improvement and peer accountability, aimed at ensuring that when the monitoring data arrives and the hydraulic models are built, the analytical foundation is robust enough to support the management plan decisions that the County of Hawaiʻi, the State, and federal funding agencies will rely on for the next 20 years. Despite the identified enhancement opportunities, adequate information exists to begin construction of reliable hydraulic models and to implement the monitoring program as designed. The commitment to document all assumptions, coordinate with the County on risk tolerance, and engage the community throughout the process — reflected throughout Task 2 — is the foundation on which a successful watershed management plan can be built.