Attachment A - NOS Modeling Requirements.pdf

Attachment A

NOAA’s National Ocean Service Modeling Requirements* Date: 6/30/20261

Table of Contents

NOAA’s National Ocean Service Modeling Requirements* 1

Introduction 1 Guiding Factors and Modeling Goals 1 Requirements for the Current State of NOS Modeling 3

1. Requirements for Accuracy, Resolution and Temporal Coverage 3

2. Requirements for Model Code 5

3. Requirements for Model Data Management 5

4. Requirements for Interoperability within Operational NOAA and NOS Modeling Ecosystems 5

5. Requirements for Geographic Coverage 6

Appendix 7

Introduction The National Ocean Service Modeling Implementation Plan (MIP) 2024 provides a detailed plan with goals, commitments, and actions to advance NOAA's ability to provide reliable coastal and ocean predictions. The plan focuses on three primary goals: engaging stakeholders to meet user needs, advancing ocean and coastal modeling through community collaboration, and delivering accurate, scalable forecasts through a national framework.

To support the gap analysis action item in the MIP, model requirements are documented here to best describe NOS’ current and future state of models to meet NOS’ mandated missions. The requirements include both quantitative (accuracy, resolution, etc.) and qualitative (functional capabilities, code requirements, etc.) to capture 1) the current state of where we are, and 2) the desired future state of NOS modeling. These requirements therefore lay the foundation for the direction of a successful NOS modeling program.

Guiding Factors and Modeling Goals The requirements outlined here are aligned with the vision, strategy, and actionable plan of NOS’s modeling program. These foundations outline that NOS’s modeling efforts will advance

1 This document is intended to be updated at least once a year, so as to dynamically adjust to emerging technologies and modeling capabilities.

https://cdn.oceanservice.noaa.gov/oceanserviceprod/tools/coastal-predictions/NOS-Modeling-Implementation-2024.pdf https://oceanservice.noaa.gov/tools/coastal-predictions/ https://cdn.oceanservice.noaa.gov/oceanserviceprod/tools/coastal-predictions/NOS-Modeling-Strategy-2023.pdf https://cdn.oceanservice.noaa.gov/oceanserviceprod/tools/coastal-predictions/NOS-Modeling-Implementation-2024.pdf its priorities of coastal resilience, conservation, and the blue economy. These efforts will provide better models for: 1) Safe, efficient marine navigation; 2) Human health protection; 3) Coastal resilience and risk reduction; 4) Improved maps and coastal management; 5) Hazardous spill response and search and rescue; and 6) Coastal and marine living resource management.

The cornerstone of providing authoritative predictions of dynamically changing ocean conditions are the hydrodynamic models which predict water levels, 3D currents, salinity and water temperature. To meet all of the above mentioned goals, these ocean models must support an integrated suite of applications, products and services. Table 1 outlines the key variables that the NOS modeling portfolio is predicting to meet its mission. The requirements for these variables were determined from user requirements, an example of which are the navigation requirements that were documented in Hess et al. (2003). These user requirements were subsequently used in the Unified Forecast System’s Coastal Application Team reports for specific applications. It is important to note that the NOS modeling ecosystem goes beyond the physical models representing ocean dynamics, to include earth geodesy models, fate/transport models (emergency response), ecological models, geospatial models, and more. Additionally, in alignment with statutory requirements in the ICOOS Act2, the IOOS Regional Associations (RAs) operate and support regionally tailored coastal ocean models. While these models are intended to specifically address regional needs, they have proven to be an asset to the NOS modeling ecosystem, complementing and expanding it by filling gaps in geographic coverage and meeting broader national stakeholder needs.

Table 1. Summary* of NOS Prediction Variables.

*Official documentation of requirements more fully outlined in Tables 2-6 in Appendix.

Variable Accuracy (Errors) Resolution Cadence and Forecast Horizon

Total Water Level Navigation:

<15 cm (0.5 ft) RMSE

Navigation (see Table 2 in Appendix for full details):

10–120m to resolve navigable waterways in ports, channels, and estuary/nearshore spatial scales.

VDatum: 30m nearshore

Coastal Shelf:

500m–3km

Human Health Protection:

100m to 1km (for estuary/nearshore to coastal) <=1 hr 3-D currents for

Navigation (see Table 2 in Appendix for full details):

Weather Scale Forecasts:

48-180 hours 180 hours to support long-range transit planning and maritime safety

Variable to meet application-dependent user requirements.

These parameters should have scientific justification showing Return on Investment for their selection.

The parameter selection should facilitate coupling with other operational NOAA

3D Currents Navigation:

<26 cm/s (0.5 kt) speed <22.5 degrees direction

+- 10% minimum of 0.1 m/sec <10 degrees for direction

3D Water Salinity Navigation:

<3.5 PSU

Pathogen: +/- 1 PSU HABs and Hypoxia: +/- 2 PSU for coastal region; +/- 5 PSU for estuaries

Hazardous Spill Response:

2 33 U.S.C. § 3601 et seq.

https://repository.library.noaa.gov/view/noaa/2460 particle trajectory applications

Sea Ice (Arctic/Alaska):

<3km in coastal zones.

Resolution sufficient to resolve narrow leads and landfast ice features

Surface currents need 10 m in rivers, small estuaries 50 m near shoreline 50 m-1km in inlets/bays, lakes <=2 km coastal ocean <=5 km in open ocean <=1 hr currents for particle trajectory applications models in the UFS.

6 hrs output frequency preferred 48 hour forecast minimum, prefer longer (72 hours for US Coast Guard)

Retrospective 3-D hindcasts:

To support ORR and NOS resilience and ecological missions with climatological normals and variability.

Variable timeframes to meet application-dependent user requirements.

10%

3D Water Temperature Navigation:

<7.7oC

Pathogen: +/- 1 ℃ HAB & Hypoxia: +/- 2 ℃

1o C

Sea Ice:

Percent coverage

Percent coverage (0–100%) to identify navigable leads (10–30%) vs. impassable zones (>70%). Skill criteria in development.

Sea Ice Thickness Thermodynamic and dynamic thickness distribution to determine ice class vessel suitability. Skill criteria in development.

Sea Ice Drift Motion vectors (speed and direction) for route optimization. Skill criteria in development.

Sea Ice Edge Accurate delineation of the Marginal Ice Zone (MIZ). Skill criteria in development.

Tidal Datums <1 cm error (model-data) <10 cm uncertainty

HAB(cells/L)/Hypoxia (Dissolved Oxygen)

Correlation >0.6 for seasonal bloom evolution;

Normalized Standard Deviation < 0.3 Spatial correlation of the blooms aligns with the model for center of mass and spatial patterns.

Requirements for the Current State of NOS Modeling The following sections outline specific criteria for inclusion in NOS’s ocean modeling portfolio.

These criteria are grouped by topic, and further considerations are also provided for the gap analysis in regards to functional capabilities.

1. Requirements for Accuracy, Resolution and Temporal Coverage

a. The ocean models will meet the skill and configuration requirements for the NOS themes of 1) Safe, efficient marine navigation; 2) Human health protection; 3) Coastal resilience and risk reduction; 4) Improved maps and coastal management; 5) Hazardous spill response and search and rescue; and 6) Coastal and marine living resource management.

i. The ocean model products will meet NOS requirements for human health protection (as captured in Table 1) and hazardous spill response and search and rescue applications (Table 5).

ii. Ocean model products should align as best possible with requirements for interagency and intra-agency downstream applications, including:

1. Search and Rescue operations of the US Coast Guard (Table 6).

2. Minimum requirements to enable short-range forecast models developed by NOAA’s Ocean Acidification Program (OAP). Ocean models deliver low-bias 3D temperature, salinity, and vertical mixing fields with resolution sufficient to be used as forcing in biogeochemical and ocean acidification models.

3. Modeling users and partners with the Navy, Department of Energy, US Army Corps of Engineers, US Geological Survey, and more.

iii. The ocean model will provide accurate water level boundary conditions for the National Water Model and accurate water levels and currents to the Nearshore Wave Prediction System of NWS.

b. The horizontal and vertical grid (if 3D) resolutions will meet user requirements as outlined in Tables 1-6. Additionally, these resolutions need to be coordinated through the UFS to align with requirements in coupled model frameworks. Model grids should be of high resolution to accurately resolve complex phenomena that could impact the key predicted variables. The targeted model resolutions, as shown below, are subject to change upon annual review of existing computational resources and modeling capabilities.

i. Navigation / ecological applications (Nearshore): High resolution (10–120m) to resolve navigable waterways in ports, channels, and estuary/nearshore spatial scales.

ii. VDatum tidal models: high-resolution coastal modeling (30 m)

iii. Coastal Shelf (Regional): Intermediate resolution (500m–3km) route optimization, upwelling/downwelling detection, ecological applications.

iv. Sea Ice (Arctic/Alaska): Resolution sufficient to resolve narrow leads and landfast ice features (ideally <3km in coastal zones).

c. The existing model forecast time periods and development of retrospective model hindcast products from NOS models will be evaluated to better meet existing NOS/NOAA office mission and external user requirements.

d. The model applications’ runtime cadence, geography, and forecast horizon are variable to meet application-dependent user requirements. These parameters should have scientific justification showing Return on Investment (ROI) for their selection. The ROI should balance considerations of physical, computational, and timing constraints based on user requirements. The parameter selection should facilitate coupling with other operational NOAA models in the UFS.

2. Requirements for Model Code

a. The core model code will be actively supported by the community. The code will be open source and accessible to the public (e.g. GitHub).

b. When put into NOAA operations, models must meet NCEP’s prediction code standards and implementation standards.

c. The code will be governed and maintained according to NAO 201-118.

3. Requirements for Model Data Management

a. The model application and code will be open source and be available to the community to improve upon and to actively participate.

i. Model output must be in a standard, community-accepted format (e.g.

NetCDF, GIS compatible, etc) and conform to NOAA data and science standards and mandates covering management, transparency, and management of data. Metadata must be CF compliant.

ii. Model data must be archived in accordance with NAO-212-15B so that it is available to NOS, the public, and other organizations for additional uses (e.g. additional data analysis, R&D, skill assessment, and hindcast reconstructions).

iii. Models used for short term forecasting (2 to 7 days) should have a station output interval of 6-minutes to align with observational water level stations.

b. Ocean models will deliver forcing fields at spatial and temporal resolutions required for offline ecological model coupling and development efforts.

c. Model outputs will align with a seamless national system for providing S-1xx (S-104 water level and S-111 currents specifications) and S-4xx suite of IHO products for navigation.

d. Model-based water levels must have output available on chart datum (Mean Lower Low Water (MLLW), local Mean Sea Level, and Lowest Astronomic Tide for coastal areas and Low Water Datum (LWD) for the Great Lakes).

4. Requirements for Interoperability within Operational NOAA and NOS Modeling

Ecosystems

a. Models deployed in NOAA operations will align with NOAA’s Unified Forecast

System (UFS). UFS is an agency-wide effort that seeks collaboration between the US government and academia, with the goal of developing a standardized model evaluation process that would improve NOAA’s ocean forecast services for the public.

b. Coastal ocean models that use hydrology as a boundary condition will link to the National Water Model by ingesting freshwater inputs overland and upstream of backwater effects.

c. The core model code will be tested using NOAA’s Unified Forecast System (UFS) Coastal Application Team (CAT) for specific applications (e.g. Navigation, Risk Reduction, Total Water Level). For applications in NOAA operations, the code should be operationally functional within the UFS-Coastal framework.

https://repository.library.noaa.gov/view/noaa/16965 https://repository.library.noaa.gov/view/noaa/16965 https://www.nco.ncep.noaa.gov/idsb/implementation_standards/ImplementationStandards.v11.0.0.pdf https://www.noaa.gov/administration/nao-201-118-software-governance-and-public-release-policy https://www.noaa.gov/sites/default/files/2025-03/NAO_212-15B_-_Data_Management_Handbook.pdf https://registry.iho.int/productspec/view.do?idx=209&product_ID=S-104&statusS=5&domainS=ALL&category=product_ID&searchValue= https://registry.iho.int/productspec/view.do?idx=210&product_ID=S-111&statusS=5&domainS=ALL&category=product_ID&searchValue= https://repository.library.noaa.gov/view/noaa/43723 https://repository.library.noaa.gov/view/noaa/44304 https://repository.library.noaa.gov/view/noaa/44304 https://repository.library.noaa.gov/view/noaa/43724

d. The entire model workflow (e.g., inputs, model run, post-processing, model output dissemination, etc.) must be completed within the required cadence or update cycle, as defined by NOS Thematic applications in the Appendix.

e. Ocean models will support interoperability standards that allow plug-and-play integration with downstream ecological models.

5. Requirements for Geographic Coverage

NOS models are configured for geographic coverage that facilitates accurate prediction in alignment with the Coast and Geodetic Act of 19473 and the Hydrographic Services Improvements Act of 19984. The model coverage overlaps with NOAA’s charting requirements. The model products should be referenced vertically in the ocean to the latest National Tidal Datum Epoch (NTDE), and to the North American-Pacific Geopotential Datum of 2022 (NAPGD2022) on land. The models’ horizontal reference system for navigation should be WGS-84.

a. The coastal ocean model will be conducive to a seamless, national modeling system and representation of the associated physical processes.

i. NOS operational models should provide complete national coverage that is to span from the coastal ocean to the inland limits of coastal influence in the full United States (CONUS and OCONUS). 10m above Mean Sea Level is the recommended inland extent to include coastal influence and facilitate linkage to the National Water Model. Lakes and major rivers with bathymetric charting/datum (including Electronic Navigational Charts) requirements (e.g. Great Lakes, Lake Champlain) and user requirements for hydrodynamic models are also included.

ii. Coverage should be composed of the fewest number of model applications possible, while still meeting accuracy, resolution, robustness (following NCEP operational requirements of 99.9% uptime), and runtime requirements, for purposes of seamless continuity of representing physical processes.

b. Inland, the NOS models need to cover at least to the end of coastal influence from tides and surge, including inundation. These combined processes affect total water levels and are needed to ensure accuracy for the NOS themes of a) safe, efficient marine navigation; b) coastal resilience and risk reduction; and c) improved maps and coastal management. As such, NOS models cover major navigable rivers of commercial interest, such as the Hudson River, Columbia River and the Mississippi River.

c. The ocean model will be coordinated with NOS’s VDatum and with NGS’s earth geodesy models (vertical land motion and gravity), frameworks (e.g. National Spatial Reference System and North-American Pacific Geopotential Datum of 2022), and data.

4 33 U.S.C. §892 et seq.

3 33 U.S.C. §883 et seq.

d. The ocean model will primarily use National Bathymetric Source (NBS) bathymetry/topography (as available), and it must allow for nontidal flooding by extending the grid onto land.

i. Models used for real-time and short-term forecasts (48 hours at a minimum) should allow for wetting and drying of intertidal areas and allow for water to inundate normally dry areas above mean higher high water.

Appendix

Table 2. Requirements for NOS Theme Safe, Efficient Marine Navigation. Marine navigation requirements from the UFS Coastal Application Team.

Need or Factor Detail/Requirement Priority variables of interest: Water levels

Surface water currents (top 25 m for navigation, top 1 m for SAR) Sea and lake ice (concentration, thickness, and velocity) Temperature and Salinity (density)

Water level accuracy 15 cm (0.5 ft) based on ~2003 estimates of pilots' needs for under keel clearance;

for time of high water and time of low water, 0.5 hr (assist in selecting port arrival/departure times)

Surface current accuracy Speed: 26 cm/sec (0.5 kt); time of max flood or ebb 30 min; for slack water times, 15 min Direction: 22.5 degrees provided current speed is not less than 26 cm/s (0.5 kt) (values based on pilots' needs for maneuvering in ports and dredged channels) For USCG SAR: 0.1 m/sec / 10 degrees

Sea and lake ice accuracy Depth/thickness 10 cm, concentration 10%, extent 10%, motion .25km/day / 10 degrees

Water density, salinity, and temperature accuracy

Desired accuracy of a forecast of a ship's draft is to the nearest 7.5 cm:

for vessel draft of 15.25 m (50 ft) (largest existing around 2003) and acceptable error in draft of 7.5 cm, acceptable error X is 3.5 psu for salinity and 7.7C for water temperature

Secondary variables of interest: Wind and atmospheric pressure, riverine, wave forcing, shorefast ice, ice pressure

Forecast frequency 6 hours

Forecast turnaround time Minimum 1 hour before forecast cycle deadline (NWS) Before start of the next model forecast cycle (NOS)

Temporal resolution of output At least hourly, optimally up to 6 minutes

Forecast range 5 to 7 days, 14 days for planning (monthly/seasonal for lake/sea ice)

Reliability 99-99.9%

Locations Coastal ocean, Great Lakes, including ports, harbors, bays, and connecting channels and rivers, and islands/atolls in the Pacific (e.g.

Hawaiian Islands and Guam) Arctic, Antarctic, Alaska, Northwest/Northeast Passages (for lake/sea ice)

Depth of currents Entire water column in order to provide currents at:

4.5 m below surface for navigation 0-1 m below surface for search and rescue

Horizontal resolution 10 m in rivers, 10s of m in shipping channels, 30 m for sea ice, 50 m-1km in inlets/bays, lakes, <=2 km around small islands, 5 km in open ocean (1 km for surface currents in EEZ) Represents these structures: levees, piers, offshore wind farms

Deterministic and/or probabilistic forecast guidance?

Definitely deterministic, interest in probabilistic

To what datum(s) do the products need to be referenced?

Vertical: chart datum (e.g. NOAA: MLLW, LWD for Great Lakes), Horizontal: WGS84

Product formats: S-100/HDF5, GRIB2, Web mapping services, GIS compatible files, NetCDF, SHEF; documentation describing files

Display systems: Have ability to overlay forcing information together on graphical interface (e.g. winds over ice); follow (S-100) portrayal rules, such as cell and other mobile devices, Portable Pilot Units (PPU), Electronic Chart Systems, and

ECDIS.

Table 3: Requirements for NOS Theme Improved Maps and Coastal Management.

Total Water Level Requirements from the UFS Coastal Application Team.

User group Need or Factor Detail/Requirement

NOAA National Centers and WFOs

Primary variables of interest

Water level Maximum inundation height and extent Significant wave height Wave direction Wave period Uncertainty estimates of these variables.

Water level accuracy WL: Bias < 0.1 m; RMSE < 0.2 m

Significant wave height accuracy

Hs: Bias < 0.3 m (or 10%); SI: < 0.3

Lead time 7-10 days 7 days for tropical

NOAA RFCs Primary variables of interest

Streamflow Precipitation data Stage Inundated area Duration of flood level Feeds into AHPS

Variable accuracy (metric)

Difference between USGS measured and model discharge.

Water level depends on rating curve. NWS maintains separate rating curves from USGS at some sites.

Categorical metric

Lead Time AHPS out to 3 to 5 days. NWM short term: 18 hours (issued hourly based on HRRR).

Medium term: 10 day (issued 4 times/day based on GFS) Long term: 30 day (issued once/day based on CFS)

FEMA, State, and Local Emergency Managers

Primary variables of interest

Peak surge, total inundation area, high water mark, Hs, time series water levels and current, morphology changes, breaching, overtopping/overwash, precip data (for operating storm drains). Capture uncertainties - what happens at days 3, 2, 1?

Variable accuracy (metric)

Aid decision on whether to order evacuations.

Lead Time Up to 7 days. Varies by EM, based on evacuation times for the particular state, county/parish, region, city, or other.

Local coastal managers

Primary variables of interest

Water level for inundation of city infrastructure (e.g storm water pipes). Local warnings of flooding. Flooding extent.

Water level accuracy Water level <0.5 foot

Lead Time 6-24 hours

USACE Districts Primary variables of interest

Peak water levels, duration of inundation, timing of flood levels (when will it reach a certain flood stage), post event recoveries, flood control. WLs relative to different vertical datums.

Water velocity and wave energy are also important quantities.

Variable accuracy (metric)

Decision on if and when to close navigation or flood gate.

Water level accuracy/uncertainty and timing window.

Lead Time 2-3 days minimum

USGS Coastal Change Viewer

Primary variables of interest

WL, Hs, Tp, Dir, tidal and longshore currents, water level relative to dune toe and crest

Water level accuracy Water level relative to dune toe/crest < 50 cm

Lead Time 6 days

USGS Storm Team

Primary variables of interest

Landfall location (peak surge and wave height), streamflow, QPF. What elevation to deploy instruments?

Variable accuracy (metric)

Go/No go decision for deployments based on storm surge uncertainty and guidance from NHC

Lead Time 4-5 days

Table 4. Requirements for NOS Theme Coastal Resilience and Risk Reduction. Risk Reduction Requirements from UFS Coastal Application Team.

Need or Factor Detail/Requirement

Water Level Accuracy Within 90% accuracy. Skill assessment calculated against: Water level observation time-series from NWLON, PORTS, and USGS gages; High Water Marks.

Hindcast Length (Period) Short term (e.g., COASTAL Act5 ~weeks) and long term (e.g., return period ~years) appropriate for the application.

Hindcast Turnaround Time Results should be available within 90 days from a named storm event.

Locations Tropical cyclones for OCONUS and islands/atolls in the Pacific (e.g., Hawaiian Islands and Guam), and US Caribbean territories (e.g., PR and USVI).

Geographic Coverage National/global Atl/GoA/PAC, Great Lakes, West Coast.

Horizontal Resolution 25 m to 120 m, depending on the elevation source data.

Resolution Possibility to produce model results to resolve inland/coastal hydraulic connections (25 m); Time step limit requirement (e.g., CFL criterion).

5 P.L. 112-141, Div. F, Title II, Subtitle B

Stability, Accuracy and Computational Efficiency

Sensitivity to bottom slope or requirement topobathy smoothing; Given required accuracy and time window for providing final results (e.g., within 90 days for COASTAL Act6); Accuracy and computational efficiency; Numerical Mixing.

Coupling Ocean, wave, inland hydrology, atmosphere, sea ice.

Data Assimilation (DA) Requirement for DA capabilities.

Code Management Community and support through GitHub or similar git platforms;

feasibility of making the model available to community developers and to be integrated into the UFS code base.

NOAA Readiness Levels Code readiness for operations, RL higher than 4/5.

License NOAA strongly encourages CC0 license for the codes developed with its support.

Spatial Reference System Vertical is NAPGD2022; Horizontal is WGS-84.

Product Formats NetCDF and CSV, depending on FEMA requirements.

Table 5. Requirements for NOS Theme Hazardous Spill Response and Search and Rescue. Office of Response and Restoration Requirements for Oceanographic Model Results

Primary variables of interest Surface currents, 3D currents

Secondary variables of interest waves, water temperature, salinity, sediment load

Currents speed accuracy +- 10% minimum of 0.1 m/sec

Currents direction accuracy 10 degrees

Waves accuracy significant wave height: 10 cm whitecapping % coverage: 10%

SST accuracy 1o C

Salinity accuracy 10%

6 P.L. 112-141, Div. F, Title II, Subtitle B

Coverage Global and Regional and Coastal (global secondary)

Locations Regional: Coastal ocean, bays, rivers,great lakes: near shorelines important.

Global: coastal and open ocean

Forecast output frequency 6 hrs

Temporal resolution of output At least 3hr, 1 hour better

Forecast range At least 48 hours, longer is better

Reliability 99%

Depth layer “Surface” currents – or top layer of model

Horizontal resolution: surface currents

10 m in rivers, small estuaries 50 m near shoreline 50 m-1km in inlets/bays, lakes <=2 km coastal ocean <=5 km in open ocean

Table 6. Requirements for NOS Theme Hazardous Spill Response and Search and Rescue. US Coast Guard Search and Rescue Requirements for Environmental Data.

Primary variables of interest wind, surface currents

Secondary variables of interest waves, sea surface temperature, air temperature, relative humidity, precipitation, visibility

Currents speed/direction accuracy 0.1 m/sec / 10 degrees

Wind speed/direction accuracy

1.0 m/sec / 10 degrees

Waves accuracy significant wave height: 10 cm whitecapping % coverage: 10%

SST accuracy 1o C

Air temperature accuracy 1o C

Relative humidity accuracy 10%

Precipitation accuracy 10%

Visibility accuracy 10%

Coverage Global and Regional

Locations Global and regional: open ocean, bays, rivers, lakes, around islands

Forecast output frequency 6 hrs

Temporal resolution of output 1 hour

Forecast range 72 hours

Reliability 99%

Wind speed & direction height @10 m

Surface currents depth layer 0-1 m

Horizontal resolution: surface currents

10 m in rivers, 50 m-1km in inlets/bays, 1 km in lakes, <=2 km around small islands <=5 km in open ocean, Horizontal resolution: winds <= 2.5 km coasts 3-12 km vast ocean

Appendix