B08_Attach_3_PMIS_236143_Final_Specifications_-_Appendix_B.pdf

National Park Service Repair of Shoreline Barriers at Smiling and Outer Coves

Wave Analysis Report

September 2020

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves, PMIS#236143, 11214635 | i

Executive summary GHD was engaged by Kaula LLC to analyze and evaluate shoreline protection measures along the Smiling Cove and Outer Cove cul-de-sac shoreline. In preparation for the development of shoreline protection measures, a Wave Analysis Report was conducted to establish the metocean design conditions for the sites. This Wave Analysis Report is part of the Supplemental Services for Task Order (TO) Number 140P2020F0163, National Park Service (NPS) Contract Number 140P8219D0001, AMME 236143.

The purpose of the Wave Analysis Report is to:

Determine wave conditions for appropriate return period events for structural survivability and retention of the shoreline.

Determine local wave climate at project sites of Smiling Cove and Outer Cove.

Assess wave parameters in the nearshore zone adjacent to the project sites.

The work carried out by GHD for this exercise included:

Data collection

Localized wave analysis

Site-specific analysis of local wave climate

A numerical model was developed for the project sites using MIKE 21 Coupled Model Flow Model. The bathymetry of Saipan and the project sites were constructed using a range of global bathymetry, satellite-derived bathymetry, and site-specific surveys. In total, eighty-eight (88) typhoons that affected Saipan in the recent 69 years between 1951 and 2019 were simulated in this exercise to establish typhoon waves.

The model was calibrated to the Saipan waverider buoy, before propagating the waves into the nearshore area. No wave measurement data inside the Tanapag Lagoon was available for calibration. The model scenarios examined included a combination of wave parameters and water levels corresponding to various return periods, as well as the projected sea level rise to the year 2070.

The model results demonstrated the effect of a fringing reef in dissipating offshore extreme wave energy. Wave height reduction by up to 91% and water level increment by up to 300% was simulated in the cases examined in this exercise. Post-analysis of the model results suggests that the dominant wave climate affecting the project sites is mostly wind-driven waves (i.e. short wave period, typically less than 8 seconds) generated inside the Lagoon, with a small percentage of the remnant offshore swell (i.e. longer wave period waves, typically longer than 8 seconds).

Wave characteristics at the project sites are mostly depth-limited, meaning the wave heights are limited by the available depth of water seaward of a location. In the scenarios examined, wind-driven waves reached a significant wave height of 2.6 feet to 4.5 feet with a peak period of 3.0 seconds to 4.2 seconds, whereas the swell mostly had a significant wave height of 0.7 to 2.9 feet with a peak period of 12.7 seconds to 15.5 seconds.

In the absence of site-specific wave measurements, the GHD modelled results were compared to historical studies for nearby locations prepared by US Army Corps of Engineers (1981) and Sea Engineering Inc. (2015). The comparison suggests the results between the studies were generally in agreement, with small discrepancies mainly contributed by the differences in water depths between the sites, the methodologies adopted by the studies, and the size off typhoon

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves, PMIS#236143, 11214635 | ii samples considered in the statistical analysis. For the purpose of designing a shoreline protection measure or a normal-importance infrastructure, the modelled results are considered acceptable in combination with the use of safety factor.

Based on a typical design scenario, the study recommended a set of design wave parameters consisted of sea wave and swell:

Parameter Smiling Cove Outer Cove

Water level incl. sea level rise to 2070 (ft. above MSL)

9.8 10.2

Wind wave Hs (ft.) 3.6 4.1 Tp (sec) 3.3 5.8

Swell wave Hs (ft.) 1.5 3.9 Tp (sec) 13.7 15.6

In addition, to mitigate the impacts of boat wake to the Smiling Cove shoreline, the study recommends a speed limit, typically < 5 knots, be implemented for the Smiling Cove navigation channel. Nevertheless, the actual speed limit should be confirmed with the relevant management authority of Smiling Cove navigation channel.

This report is subject to, and must be read in conjunction with, the limitations set out in Section

1.4 and the assumptions and qualifications contained throughout the Report.

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves, PMIS#236143, 11214635 | i

Table of contents

1. Introduction

1.1 Project background

1.2 Purpose

1.3 Scope of work

1.4 Clarifications / Exclusions

1.5 Disclaimers

2. Physical data

2.1 Datum

2.2 Bathymetry model

2.3 Wind climate

2.4 Buoy wave data

2.5 WaveWatch III hindcast data

2.6 Tidal planes

2.7 Sea level rise

3. Numerical modelling

3.1 Regional model

3.2 Local model

3.3 Model results

3.4 Impacts from propeller wash

4. Conclusions and recommendations

4.1 Conclusions

4.2 Recommendations

5. References

Table index Table 1 Wind speed exceedance fractile values

Table 2 Critical typhoons

Table 3 Tide elevations of Saipan, station 1633227

Table 4 Projected SLR for intermediate and extreme emission scenarios

Table 5 Water level for each return period at the buoy location

Table 6 Hs for each return period at the buoy location

Table 7 Extraction points depth and coordinates

Table 8 Summary of offshore Hs and water level combination cases used for nearshore modelling

Table 9 Water level (ft. above MSL) and current speed (kt) at each extraction point

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves, PMIS#236143, 11214635 | ii

Table 10 Significant wave height, Hs (ft.) at each extraction point

Table 11 Maximum wave height, Hmax (ft.) at each extraction point

Table 12 Peak wave period, Tp (sec) at each extraction point

Table 13 Model results - Case 4 with 2070 SLR for intermediate emission

Table 14 Recommended design wave parameters

Figure index Figure 1 Outer Cover and Smiling Cove Project sites (Google Earth)

Figure 2 Offshore bathymetry model using GEBCO data

Figure 3 Nearshore bathymetry model

Figure 4 Processed monthly wind rose based on CFS hourly wind data (1979 – 2019)

Figure 5 Processed annual wind rose based on CFS hourly wind data (1979 – 2019)

Figure 6 Wind speed exceedance probability based on CFS hourly wind data (1979 – 2019)

Figure 7 Tracks of identified critical typhoons

Figure 8 Modelling flowchart

Figure 9 Regional domain showing Saipan and Tinian islands

Figure 10 Model results and Station 52211 buoy data comparison for year 2016 – (top:

Significant wave height; bottom: Peak wave period)

Figure 11 Maximum simulated water level for each typhoon extracted at the buoy location

Figure 12 Maximum simulated HS for each typhoon extracted at the buoy location

Figure 13 Hs vs. Tp and Hs vs. water level

Figure 14 Return period curve for water level (top pan) and Hs (bottom pan)

Figure 15 Wave breaking coefficient map

Figure 16 Local domain with indicated observation points and transects

Figure 17 Wave breaking at the reef

Figure 18 Hs (top pan) and water level profile (bottom pan) plots along T1

Figure 19 Hs (top pan) and water level profile (bottom pan) plots along T2

Figure 20 Hs (top pan) and water level profile (bottom pan) plots along T3

Figure 21 Separation of sea waves and swell in Case 4 along T2

Appendices Appendix A Joint probability tables of Station 52211 Saipan buoy data for year 2016

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves, PMIS#236143, 11214635 | iii

Acronyms Acronym Meaning ARI Averaged Recurrence Interval CFS Climate Forecast Model System DHI Dansk Hydraulisk Institut, the developer of the MIKE model EVA Extreme value analysis GEBCO General Bathymetric Chart of the Oceans HD Hydrodynamic, a modelling module by DHI MIKE Hmax Maximum wave height Hs Significant wave height JMA-RSMC Japan Meteorological Agency Regional Specialized Meteorological Center MSL Mean sea level NCEP National Centers for Environmental Prediction NOAA National Oceanic and Atmospheric Administration SW Spectral Wave, a modelling module by DHI MIKE Tp Peak wave period WWIII Wave Watch III, third generation wave model developed by NOAA/NCEP

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 4

1. Introduction

1.1 Project background

GHD was engaged by Kaula LLC to analyze and evaluate shoreline protection measures along the Smiling Cove and Outer Cove cul-de-sac shoreline (the project sites). In preparation for the development of shoreline protection measures, a Wave Analysis Report was conducted to establish the metocean design conditions. The project sites are shown in Figure 1.

Figure 1 Outer Cover and Smiling Cove Project sites (Google Earth)

1.2 Purpose

This Wave Analysis Report is part of the Supplemental Services for Task Order (TO) Number 140P2020F0163, National Park Service (NPS) Contract Number 140P8219D0004, AMME 236143.

The purpose of the Wave Analysis Report is to develop design wave conditions, and any potential prop wash impacts, for determining the design criteria for the proposed shoreline protection along the Smiling Cove entrance channel and Outer Cove shoreline.

1.3 Scope of work

The scope of work carried out by GHD under this contract included:

Data collection

Localized wave analysis to determine:

– design wave conditions

– any potential prop wash impacts

– wave conditions for appropriate return period events for structural survivability and retention of the shoreline

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 5 design criteria for the proposed shoreline protection along the Smiling Cove entrance channel and Outer Cove shoreline.

Site-specific analysis of local wave climate to determine:

– wave conditions for appropriate return period events for structural survivability and retention of the shoreline

– locally wind-driven waves and offshore waves approaching the project site of Smiling Cove and Outer Cove

– wave parameters in the nearshore zone adjacent to the project site

1.4 Clarifications / Exclusions

Wind data used in the analysis for the ambient condition was sourced from the Climate Forecast Model System – CFS reanalysis, instead of actual field measurement due to unavailability of such data. For the purpose of this exercise and for deriving the design condition for shoreline protection structures, the data from the CFS reanalysis is deemed adequate as the governing force for the site is from typhoon events.

Wind data used in the analysis for the extreme condition was sourced from the Japan Meteorological Agency which serves as the Regional Specialized Meteorological Center for the western North Pacific basin.

Wave data from Station 52211 - Tanapag, Saipan, CNMI was used to evaluate and calibrate the model performance for offshore wave conditions. Model calibration was not performed for nearshore results (i.e. inside the Tanapag Lagoon) due to the lack of site measurement data. Researches, literature, and papers relevant to understand and model the behavior of wave transformation over fringing reefs were referenced and considered in setting up the numerical model.

1.5 Disclaimers

This report has been prepared by GHD for National Park Service and may only be used and relied on by National Park Service for the purpose agreed between GHD and the National Park Service as set out in Section 1.2 of this report.

GHD otherwise disclaims responsibility to any person other than National Park Service arising in connection with this report. GHD also excludes implied warranties and conditions, to the extent legally permissible.

The services undertaken by GHD in connection with preparing this report were limited to those specifically detailed in the report and are subject to the scope limitations set out in the report.

The opinions, conclusions and any recommendations in this report are based on conditions encountered and information reviewed at the date of preparation of the report. GHD has no responsibility or obligation to update this report to account for events or changes occurring subsequent to the date that the report was prepared.

The opinions, conclusions and any recommendations in this report are based on assumptions made by GHD described in this report. GHD disclaims liability arising from any of the assumptions being incorrect.

GHD has prepared this report on the basis of information provided by National Park Service and others who provided information to GHD (including Government authorities), which GHD has not independently verified or checked beyond the agreed scope of work. GHD does not accept liability in connection with such unverified information, including errors and omissions in the report which were caused by errors or omissions in that information.

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 6

GHD has not been involved in the preparation and has no contribution to, or review of the document other than those prepared by GHD. GHD shall not be liable to any person for any error in, omission from, or false or misleading statement in, any other part of the document not prepared by GHD.

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 7

2. Physical data

2.1 Datum

The following datum was adopted in the modelling work:

The horizontal datum for the numerical model: UTM 55N

The vertical datum is referred from Mean Sea Level (MSL).

2.2 Bathymetry model

Two scales of bathymetry models were used in the modelling exercise:

Offshore bathymetry was modelled using the General Bathymetric Chart of the Oceans (GEBCO) gridded bathymetric data set. The GEBCO_2020 grid is a global terrain model for ocean and land at 15 arc-second intervals.

Nearshore bathymetry including the fringing reef and lagoon was based on the 5 meter gridded data from the Pacific Islands Ocean Observing System (PacIOOS). For the project sites, a 10-feet Saipan Harbor hydrographic survey by Sea Engineering Inc.

(2017) and a site-specific hydrographic survey of 5 feet resolution prepared by Dueñas, Camacho & Associates, Inc. (DCA, 2020) were adopted.

Figure 2 and Figure 3 shows the offshore and nearshore bathymetry models respectively.

Figure 2 Offshore bathymetry model using GEBCO data

PROJECT SITES

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 8

Figure 3 Nearshore bathymetry model

2.3 Wind climate

2.3.1 Ambient wind

Wind data was sourced from the Climate Forecast Model System – CFS reanalysis, which is managed and operated by the United States of America, National Oceanic and Atmospheric Administration (NOAA).

Analysis of ambient wind condition was extracted at a point near the project sites (15°13’27.8” N, 145°43’13.2” E), as shown in Figure 3, for the years 1979 - 2019.

The ambient wind climate for this dataset is summarized in monthly and annual wind roses in Figure 4 and Figure 5. The exceedance probability plot is shown in Figure 6.

The ambient wind climate can be characterized as follows:

95% of hourly wind speeds are less than 21 knots (kt) (36 feet per second [ft./s])

Occurrence of wind speeds greater than 26 knots (44.5 ft./s) are rare (~1%), and may be associated with typhoon events

Dominant wind pattern is the easterly trade wind, with directions following a seasonal cycle:

– Seasonal trade winds from the northeast from November to March (colored in red), and

– Easterly winds from May to October (colored in blue)

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 9

Figure 4 Processed monthly wind rose based on CFS hourly wind data (1979

– 2019)

JAN FEB MAR

APR MAY JUN

JUL AUG SEP

OCT NOV DEC

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 10

Figure 5 Processed annual wind rose based on CFS hourly wind data (1979 – 2019)

The wind speed values at selected fractiles presented in Figure 6 are summarized in Table 1.

These values represent the percentage of ambient wind exceeded at these fractiles. It is expected that 11 kt winds exceed 95% of the ambient wind experienced at the site. Conversely, ambient winds greater than 11 kt have a 5% probability.

Figure 6 Wind speed exceedance probability based on CFS hourly wind data

(1979 – 2019)

Table 1 Wind speed exceedance fractile values

Fractile Wind Speed (kt)

25.0 % 4.95

50.0 % 6.6

75.0 % 8.3

95.0 % 11.0

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 11

2.3.2 Extreme wind

Extreme wind events in Saipan are generally associated with typhoons. Typhoon data was sourced from the Japan Meteorological Agency Regional Specialized Meteorological Center (JMA-RSMC Tokyo).

Figure 7 shows the historical typhoon tracks that passed near the project site from 1951 to 2019, as summarized in Table 2. In total, 88 historical typhoons were identified based on the following selection criteria, including:

Their tracked passed the project sites within 185 mile (mi) (~300 kilometer [km]) radius

The typhoons have a linearly varying pressure gradient from ambient atmospheric pressure to 950 hectopascals (hPa) at the 185 mi. radius.

Figure 7 Tracks of identified critical typhoons

Table 2 Critical typhoons

Year International Name

Minimum Pressure (hPa)

Year International Name

Minimum Pressure (hPa)

1951 Thelma 950 1989 Winona 1000 1953 Rita 940 1989 Andy 920 1953 Doris 935 1989 Colleen 950 1954 No name 995 1989 Forrest 940 1954 Lorna 950 1990 Koryn 980 1957 Trix 950 1990 Russ 915 1957 Judy 960 1991 Mireille 925 1957 Lola 900 1991 Seth 925 1958 Tess 1000 1992 Omar 920 1958 Olga 950 1992 Gay 900 1962 Marge 1002 1992 Hunt 940 1962 Ruth 915 1993 Steve 980 1962 Emma 890 1994 Verne 950

185 mi. radius

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 12

Year International Name

Minimum Pressure (hPa)

Year International Name

Minimum Pressure (hPa)

1962 Karen 900 1994 Wilda 935 1962 Nadine 974 1994 Zelda 910 1963 Olive 920 1994 Bobbie 968 1963 Susan 935 1996 Yates 935 1965 Bess 900 1997 Winnie 915 1965 Carmen 914 1997 Joan 905 1967 Therese 990 1997 Keith 910 1967 Gilda 910 1997 Paka 920 1967 Harriet 950 2000 Saomai 925 1968 Jean 935 2001 Faxai 915 1968 Wendy 920 2002 Chataan 930 1968 Agnes 900 2002 Pongsona 940 1968 Irma 955 2004 Tingting 955 1968 Kit 960 2004 Chaba 910 1969 Elsie 895 2005 Nabi 925 1970 Patsy 910 2007 Kong-rey 960 1972 Betty 910 2009 Choi-wan 915 1974 Babe 985 2009 Melor 910 1974 Carla 965 2014 Vongfong 900 1974 Polly 950 2015 Dolphin 925 1976 Pamela 920 2015 Nangka 925 1976 Therese 905 2015 Soudelor 900 1976 Billie 915 2015 Goni 930 1980 Dinah 950 2015 Dujuan 925 1981 Gerald 980 2015 Koppu 925 1981 Hazen 955 2015 Champi 930 1984 Thad 935 2018 Jebi 915 1986 Kim 905 2018 Mangkhut 905 1987 Wynne 920 2018 Yutu 900 1987 Lynn 920 2019 Hagibis 915 1988 Roy 940 2019 Bualoi 935

2.4 Buoy wave data

Wave measurement data was retrieved from Station 52211 Tanapag, Saipan waverider buoy at 15°16'5" N 145°39'44" E, previously shown in Figure 3. The water depth at its location is approximately 1,600 ft. Joint probability tables for joint occurences of wave heights and wave periods, wave heights and wave directions are presented in Appendix A.

The following were observed from the 2016 buoy data:

95.6% of the offshore significant wave height Hs was less than 10 ft.

The top three wave directions recorded were NE (36.5%), NNE (20.7%), and ENE (17.5%). These three directions accounted for 75% of the recorded waves in 2016.

97.4% of the offshore waves had a peak wave period between 6 and 15 seconds, with 60.8% falling within the 9 to 12 seconds range. This implies that a large portion of the wave climate in the region is dominated by swell waves.

This dataset will be used to compare against model results in section 3.1.1.

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 13

2.5 WaveWatch III hindcast data

The boundary conditions implemented in the numerical model were based on the data extracted from the NOAA WaveWatch III model (WWIII).

The WWIII model is a third generation global wave model that benchmarks its results with buoy measurements located in the Oceans. The model updates its database four times a day, each run starts with 9-, 6- and 3-hour hindcasts and produces forecasts of every 3 hours from the initial time out to 180 hours. The model solves the random phase spectral action density balance equation for wavenumber-direction spectra.

2.6 Tidal planes

Tide elevations of Saipan are taken from the NOAA tides and currents inquiry page and National Geodetic Survey for Station 1633227, Saipan[1][2], and is summarized in Table 3.

Table 3 Tide elevations of Saipan, station 1633227

Tidal plane Elevation (ft.

MLLW)

Elevation

(ft. MSL)

Elevation (ft.

NMVD03[3])

Elevation (ft., NGS Benchmark

GRPN-1)

Highest Astronomical Tide

(HAT)

2.49 1.20 -4.20 -3.56

Mean Higher High Water

(MHHW)

2.20 0.91 -4.49 -3.85

Mean High Water (MHW) 2.07 0.78 -4.62 -3.98 Mean Seal Level (MSL) 1.29 0.00 -5.40 -4.76 Mean Low Water (MLW) 0.61 -0.68 -6.08 -5.44 Mean Lower Low Water (MLLW) 0.00 -1.29 -6.69 -6.05 Lowest Astronomical Tide (LAT) -0.96 -2.25 -7.65 -7.01

2.7 Sea level rise

Similar to all coastlines around the world, the coastline of Saipan is subject to the impacts of sea level rise (SLR) and storm surge. In a paper published by NOAA (Sweet et al, 2017) as referenced in Sea Engineering Inc (2018) and the Sea Level Rise Viewer compiled by NOAA[4], the SLR at Saipan relative to MHHW is projected to reach the magnitudes presented in Table 4.

Table 4 Projected SLR for intermediate and extreme emission scenarios

Year Projected SLR for intermediate emission

Projected SLR for extreme emission

2030 0.8 ft. 1.2 ft.

2050 1.9 ft. 2.8 ft.

2070 3.4 ft. 5.3 ft.

2100 6.3 ft. 10.4 ft.

1 https://tidesandcurrents.noaa.gov/datums.html?datum=MLLW&units=1&epoch=0&id=1633227&name=Saipa%20n&state= 2 https://www.ngs.noaa.gov/datums/vertical/northern-marianas-vertical-datum-2003.shtml 3 North Mariana Vertical Datum 2003 4 https://coast.noaa.gov/slr/#/layer/sce/0/16221669.400715254/1714730.2011476331/14/satellite/106/0.8/2070/extreme/mid Accretion https://coast.noaa.gov/slr/#/layer/sce/0/16221669.400715254/1714730.2011476331/14/satellite/106/0.8/2070/extreme/mid

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 14

3. Numerical modelling The modelling approach is outlined in Figure 8. A two-level domain scheme, comprised of regional and local domains, was implemented to establish the offshore and nearshore wave and water level conditions[5].

Figure 8 Modelling flowchart

3.1 Regional model

The regional domain covers the open ocean and islands of Saipan and Tinian, as shown in Figure 9. Other smaller, farther islands were excluded in the modeling since they would have a negligible effect on the wave climate at the sites.

5 The MIKE 21 Coupled Model Flow Model modelling software was used in the modelling exercise. The model is a dynamic modelling system for applications within coastal, estuarine and river environments. Its primary components, the Hydrodynamic

(HD) and the Spectral Wave (SW) modules are capable of simulating the mutual interaction between waves and currents through dynamic coupling (DHI, 2017).

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 15

Figure 9 Regional domain showing Saipan and Tinian islands

3.1.1 Model evaluation

Ambient wave conditions at regional scale were simulated for the year 2016, nominated as a representative year since no severe typhoon event was recorded during this year as evident in Table 2. CFS wind data and WWIII hindcast wave parameters for the same period were used as the boundary conditions.

Figure 10 shows the comparison between the model results (red line) and the measured data available from Station 52211 Tanapag, Saipan waverider buoy data (blue line) for both significant wave height (Hs) and peak wave period (Tp). The fitted data has an mean absolute error of 1.1 ft., with root mean square error of 1.3 ft., and index of agreement equal to 0.9, from a scale of 0 (no agreement) to 1 (perfect match). From the comparison, it can be seen that the observed data was reproduced in the MIKE21 model with overall consistent trend and limited discrepancy.

The small discrepancy is most likely due to the coarse offshore bathymetry resolution and the inaccuracies from the boundary conditions extracted from CFS winds and WWIII dataset; in addition, these boundary conditions were generated by hindcast models that may deviate from the actual, dynamic, and complex metocean conditions. Nevertheless, GHD deemed the model setup including the model mesh and bathymetry adequate for capturing the offshore metocean characteristics.

Saipan

Tinian

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 16

Figure 10 Model results and Station 52211 buoy data comparison for year 2016 – (top: Significant wave height; bottom: Peak wave period)

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 17

3.1.2 Typhoon modelling

The MIKE 21 Coupled Model Flow Model simulated storm tides and extreme waves generated by the 88 typhoons listed in Table 2.

Figure 11 shows the peak water level for each simulated typhoon extracted at the location of the Saipan waverider buoy. The highest water level from all typhoons was found to be caused by 1968 Typhoon Jean, which reached a maximum water level of 3.2 ft. MSL. Typhoon Jean, at its peak, was classified as a Category 3 hurricane with a 1-minute sustained wind speed of 115 miles per hour (mph) and a minimum central pressure of 950 hPa.

Figure 12 shows the significant wave height (Hs) for each typhoon extracted at the location of the buoy. The resulting Hs is a combination of wind waves and swell waves. The highest typhoon wave from all typhoons was found to be caused by 1976 Typhoon Therese with Hs of 57 ft., primarily composed of swell waves components. Typhoon Therese, at its peak, was classified as a Category 4 hurricane with a 1-minute sustained wind speed of 155 mph and a minimum central pressure of 905 hPa.

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 18

Figure 11 Maximum simulated water level for each typhoon extracted at the buoy location

Figure 12 Maximum simulated HS for each typhoon extracted at the buoy location

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 19

Figure 13 shows the relationship of Hs to Tp and water level at the buoy extraction point. Each point in the scatter plot corresponds to the peak of every critical typhoon. The trend in the top plot demonstrates a positive dependence between Hs and Tp, suggesting that the larger the wave height, the likely the longer the wave period, and vice versa. Furthermore, the majority periods of the peak waves are more than 10 seconds, signifying that offshore wave conditions during typhoons are mostly swell dominated.

The bottom plot trend also demonstrates some dependence of water level with Hs, but the correlation is not as strong as between Hs and Tp. Generally, a larger wave height can be expected to occur together with a higher water level.

Figure 13 Hs vs. Tp and Hs vs. water level

3.1.3 Extreme value analysis

Extreme value analysis (EVA) to estimate the wave heights and water levels corresponding to various Averaged Recurrence Interval (ARI) was undertaken by fitting a Weibull distribution to the extreme wave and water level values. Figure 14 shows the return period plots for water level and Hs at the offshore extraction point at the waverider buoy with a 95% confidence band. This means that from the given data, 95% of possible scenarios would be inside this upper and lower limit. Given the typhoon data is only available from 1951 to 2019 (a total of 69 years), the values presented for the 100-yr ARI return period are extrapolated.

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 20

Figure 14 Return period curve for water level (top pan) and Hs (bottom pan)

Table 5 and Table 6 summarize the above figures into the tables. The Hs values presented in this table are a combination of both wind wave and swell components. Generally, higher waves occur less often and have a higher return period.

Table 5 Water level for each return period at the buoy location

Return period

(yr. ARI)

Water level (ft.

from MSL)

2 1.4±0.08 5 1.8±0.17 10 2.2±0.27 20 2.5±0.4 50 2.9±0.6 100 3.1±0.8

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 21

Table 6 Hs for each return period at the buoy location

Return period

(yr. ARI)

Sig. wave height Hs (ft.)

2 26.7±2.4 5 36.2±3.6 10 42.3±4.9 20 48±6.3 50 54.9±8.4 100 59.8±10

3.2 Local model

3.2.1 Simple wave hindcasting

Since there is no site measurement data available to calibrate the nearshore model, a simple wave hindcast based on the JONSWAP parametric method as described in Kamphuis (2000) was performed to estimate the wind-generated waves inside the reef lagoon during a typhoon event[6]. The results of the hindcasting were compared with the model results in the following sections.

The JONSWAP parametric method determines the Hs and Tp from the following parameters:

Fetch (F): fetch is the distance the wind blows over the water to generate the waves. For a reef lagoon system, where most of the offshore waves break at the reef edge during a typhoon event, the fetch can be estimated as the distance from the project sites to the reef boundary. The available fetch from the project sites to the reef edge is approximately

3.5 km.

Storm duration (t): analysis of several typhoons that passed the project sites indicated that the typhoons maintained their peak wind speed blowing to the project sites for no longer than 2 hours.

Depth of water in the fetch (d): averaged to be 14 ft.

Using a 1-minute averaged wind speed of 150 mph, corresponding to a Category 5 hurricane, and an average depth of 14 ft. near the project sites, the estimated wind-generated Hs and Tp inside the lagoon were in the range of 5.2-5.9 ft. and 3.7-4.9 sec. Allowing the wind to blow for a longer duration did not result in higher waves because the growth of the waves inside the lagoon was limited by both the available fetch length and available depth at the project sites.

It is noted that the parametric hindcasting did not account for complex wave transformation processes happening inside the reef lagoon nor for the varying bathymetry as the wave propagates. As such, the results of the hindcasting serve only as a first-order approximation.

3.2.2 Model setup

The local domain, as shown in Figure 16, allowed for detailed modelling of the project sites. The mesh has increasing resolution from offshore to nearshore, including atop the fringing reef, to capture wave transformation and hydrodynamics as the water depth decreases.

A varying wave breaking coefficient, γ, which determines wave breaking as a ratio of wave height to water depth, was implemented in the model. Yao, Huang, Monismith, & Lo (2012) showed that the relative reef-flat submergence is an important factor to characterize most wave-

6 Site measurement carried out by Cai, Chen, Liu, & Sun (2019) showed that waves generated inside the lagoon were generally in good agreement with the fetch-limited wave theory.

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 22 breaking features such as the breaker type and location, the surf-zone width, and the incipient breaker depth index in a fringing reef setting. In addition, based on a previous project carried out by GHD on a similar fringing reef environment, the γ values of 0.78 over the reef and 0.55 in the lagoon showed a good agreement with site observations. As such, the γ values used in this study varied from 0.8 in offshore areas, 0.78 at the reef edge, and 0.55 within the reef. Figure 15 shows the gamma values assigned to the model.

Figure 15 Wave breaking coefficient map

Model results were extracted at nine extraction points placed within the project sites; with five points along Smiling Cove, labelled as S1-S5, and four points along Outer Cove, labelled as O1- O4. The coordinates and depth of the extraction points are summarized in Table 7.

Table 7 Extraction points depth and coordinates

Points Depth

(ft. from MSL)

Easting Northing

S5 -7.60 362871 1683264 S1 -7.14 362909 1683073 S2 -8.73 362921 1683007 S3 -7.17 362939 1682942 S4 -7.46 362941 1682889 O1 -1.38 363218 1683270 O2 -2.26 363231 1683232 O3 -1.90 363236 1683197 O4 -1.69 363240 1683160

Two transects, T1 and T2, were used to analyze the wave transformation and hydrodynamics over the offshore fringing reefs. Wave breaking is particularly important for this scenario since the abrupt change in bathymetry at the reef would condense the wave energy near the surface leading to wave breaking. A nearshore transect, T3, was placed along the channel at Smiling

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 23

Cove to investigate the wave transformation in the channel such as wave reflection and diffraction around obstacles. Figure 16 illustrates the locations of the transects.

Figure 16 Local domain with indicated observation points and transects

3.2.3 Summary of model cases

The results of the EVA were used to establish the wave and water level combinations to be used as the boundary conditions for the local domain model.

In lieu of an extensive joint probability analysis through Monte Carlo simulation process and that adopting a 100 yr ARI water level with a 100 yr ARI wave height may be conservative, based on our previous modelling experience in cyclone-prone areas GHD nominates using the water level

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 24 and wave height combinations summarized in Table 8 to evaluate the effects of difference water levels and extreme wave heights.

Table 8 Summary of offshore Hs and water level combination cases used for nearshore modelling

Case Water level RP (yr. ARI)

Sig. wave height RP (yr. ARI)

Water level (ft.

from MSL)

Sig. wave height Hs (ft.)

Peak wave period Tp (sec)

1 100 20 3.1 48.0 15.2 2 50 20 2.9 48.0 15.2 3 20 50 2.5 54.9 16.3 4 20 100 2.5 59.8 17.0

During actual typhoons, waves may approach from various directions depending on the track of the typhoon generating the waves. For this exercise, a sensitivity check was performed to determine the direction of wave propagation that generated the largest wave height at the project sites. A unit, monochromatic wave was propagated from the north boundary with varying directions in north-northeast (NNE), north (N), and north-northwest (NNW). Based on the sensitivity check, the critical direction was found to be from the NNW.

3.3 Model results

3.3.1 Effect of fringing reef

Figure 17 illustrates the location of wave breaking simulated in the model. As expected, the breaking mainly occurred across the fringing reef. Figure 18 and Figure 19 show the results for Cases 1 to 4 along the offshore transects T1 and T2 respectively, with the significant wave height and water level from mean sea level plotted together with the bathymetry.

The effect of the fringing reef is evident in the above profile plots. Heights of the incoming waves decreased dramatically at the reef edge where the seabed slope is steep and the available water depth is limited. Due to the presence of the breaking waves and wave setup, the mean water level is increased while the wave heights remained similar within the reef between the cases. For the cases investigated, the reduction in significant wave height was seen by up to 91%, from 62 ft[7]. at the offshore side of the fringing reef to 6 ft. inside the Lagoon after breaking; the increment in water level reached up to 300%, from 2.3 ft. at the offshore side of the fringing reef to 6.9 ft. inside the Lagoon.

7 The 100 yr ARI wave height adopted at the model boundary is 59.8 ft. As the wave propagated nearshore, shoaling effects increased the wave height.

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 25

Figure 17 Wave breaking at the reef

PROJECT SITE

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 26

Figure 18 Hs (top pan) and water level profile (bottom pan) plots along T1

Edge of fringing reef

Saipan Harbor channel

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 27

Figure 19 Hs (top pan) and water level profile (bottom pan) plots along T2

Edge of fringing reef

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 28

Secondary wave breaking further along the transect T1 was also simulated due to decrease in water depth after the transect passes the dredged Saipan Harbour navigation channel. As for T2, both the water level and significant wave height remained relatively stable beyond the initial breaking zone at the fringing reef.

Transect T3 is taken along the Smiling Cove channel, as seen in Figure 20. As the water depth started to become shallower at the channel entrance, the incident waves broke further. The model suggests the significant wave height decreased by about 5 ft. along T3 between the entrance and the end of the transect. The water level, however, appeared to be mostly constant along the channel.

3.3.2 Sea and swell separation

For all the cases investigated above, Case 4 consistently presents the worst-case scenario. A closer inspection of the Case 4 wave breaking at the reef edge was undertaken by separating the wind sea and swell components of the wave as it propagates along T2, as shown in Figure 21.

It is seen that the wave climate was mostly swell-dominant in offshore before started breaking at the reef edge. Inside the reef lagoon, the majority of the swell wave energy was dissipated by the reef subsequently, wind waves became the dominant wave. A similar phenomenon was observed for all other cases.

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 29

Figure 20 Hs (top pan) and water level profile (bottom pan) plots along T3

Smiling Cove channel entrance

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 30

Figure 21 Separation of sea waves and swell in Case 4 along T2

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 31

3.3.3 Model outputs

Table 9 summarizes the water level and current speed extracted at the extraction points. Water level simulated in Case 4 was the highest amongst other cases: 7.5 ft. above MSL at Smiling Cove and 7.9 ft. above MSL at Outer Cove. Current speeds are generally constant at each project sites. For the extraction points at Outer Cove, the most exposed extraction point O1 and O4 exhibit the highest current speed.

Table 9 Water level (ft. above MSL) and current speed (kt) at each extraction point

Point Case 1 Case 2 Case 3 Case 4 Water Level

Current Speed

Water Level

Current Speed

Water Level

Current Speed

Water Level

Current Speed

S5 6.6 0.8 6.2 0.8 6.9 0.8 7.2 0.8 S1 6.9 0.8 6.6 0.8 6.9 0.8 7.5 0.8 S2 6.9 0.8 6.6 0.8 6.9 0.8 7.5 0.8 S3 6.9 0.8 6.6 0.8 6.9 0.8 7.5 0.8 S4 6.9 0.8 6.6 0.8 6.9 0.8 7.5 0.8 O1 6.9 0.4 6.6 0.4 7.2 0.4 7.9 0.4 O2 6.9 0.2 6.6 0.2 7.2 0.2 7.9 0.2 O3 6.9 0.2 6.9 0.2 7.2 0.2 7.9 0.2 O4 7.2 0.4 6.9 0.4 7.2 0.4 7.9 0.4

Table 10 to Table 12 summarize the Hs, Hmax, and Tp extracted at the extraction points. For these three parameters, the values are presented for both the wind sea and swell components.

Case 4 simulated the highest wave Hs and Hmax amongst other cases. The wave heights are seen to decrease along the Smiling Cove channel from S1 to S4 and along the coast of Outer Cove from O1 to O4 as the water depth decreases (refer to discussion in Section 3.3.1).

As shown in Table 10 and Table 11, the wind wave components are larger than the swell wave components for all extraction points across all cases investigated. This is especially true for the points inside the Smiling Cove channel (S1-S4) where the swell wave heights are substantially lower than the wind-wave heights.

Table 10 Significant wave height, Hs (ft.) at each extraction point

Point Case 1 Case 2 Case 3 Case 4

Wind wave

Swell wave

Wind wave

Swell wave

Wind wave

Swell wave

Wind wave

Swell wave

S5 4.4 1.9 4.3 1.8 4.4 2.0 4.5 2.2

S1 3.1 0.8 3.0 0.7 3.1 0.9 3.3 1.0

S2 2.9 0.7 2.9 0.6 3.0 0.7 3.1 0.8

S3 2.8 0.6 2.7 0.5 2.8 0.6 3.0 0.7

S4 2.7 0.6 2.6 0.5 2.7 0.7 2.9 0.8

O1 3.5 2.4 3.4 2.3 3.5 2.6 3.6 2.9

O2 3.3 2.0 3.2 1.9 3.3 2.1 3.4 2.3

O3 2.8 1.6 2.7 1.5 2.8 1.7 2.9 1.9

O4 2.4 1.4 2.4 1.3 2.5 1.5 2.6 1.7

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 32

Table 11 Maximum wave height, Hmax (ft.) at each extraction point

Point Case 1 Case 2 Case 3 Case 4 Wind wave

Swell wave

Wind wave

Swell wave

Wind wave

Swell wave

Wind wave

Swell wave

S5 6.7 3.5 6.6 3.3 6.6 3.7 6.8 4.1 S1 5.9 1.4 5.7 1.3 6.0 1.6 6.3 1.8 S2 5.8 1.2 5.6 1.1 5.9 1.3 6.2 1.5 S3 5.3 1.0 5.2 1.0 5.4 1.1 5.7 1.3 S4 5.4 1.1 5.3 1.0 5.5 1.2 5.8 1.4 O1 5.4 4.5 5.2 4.3 5.4 4.8 5.6 5.2 O2 5.5 3.6 5.4 3.5 5.5 3.9 5.7 4.2 O3 5.4 3.0 5.3 2.9 5.4 3.2 5.7 3.5 O4 4.9 2.6 4.8 2.5 5.0 2.8 5.2 3.1

Table 12 Peak wave period, Tp (sec) at each extraction point

Point Case 1 Case 2 Case 3 Case 4

Wind wave

Swell wave

Wind wave

Swell wave

Wind wave

Swell wave

Wind wave

Swell wave

S5 3.3 12.0 3.3 11.8 3.3 12.1 3.3 12.7

S1 5.9 12.0 2.9 11.9 3.0 12.6 3.0 13.2

S2 5.8 11.9 2.9 11.9 2.9 12.6 3.0 13.2

S3 5.3 12.0 2.8 12.0 2.9 12.8 3.0 13.3

S4 5.4 12.0 3.0 11.9 3.0 12.7 3.1 13.3

O1 5.4 14.1 5.1 14.2 5.1 15.0 4.2 15.4

O2 5.5 13.8 3.6 13.8 3.4 14.8 3.4 15.3

O3 5.4 14.1 3.6 14.0 3.5 15.1 3.5 15.5

O4 4.9 14.0 3.7 14.0 3.6 14.9 3.5 15.4

3.3.4 Results comparison

Hindcast values obtained via fetch-limited wave estimate were compared with model results at S5, whose location, depth, and exposure are similar to the parameters adopted in the hindcast.

Hindcast values of 5.2-5.9 ft. Hs and 3.7-4.9 seconds Tp are in reasonable agreement with the results at S5 of 4.5 ft. Hs. and 3.3 seconds Tp for Case 4. In addition, the model results were compared to the earlier studies for projects nearby Smiling and Outer Cove areas prepared by U.S. Army Corps of Engineering (1981) whom adopted the similar fetch-limited wave estimate and Sea Engineering Inc. (2015) whom deployed another suite of numerical model (SWAN).

In general, the GHD model results are in agreement with the other studies, with small discrepancies in the GHD results. The differences in wave parameters between the studies mainly originated from the variance in water depths at the respective study sites (the three studies were for different but nearby locations), and that the complex coastal processes such as wave spreading, breaking, refraction, and diffraction were considered in the GHD model whereas such processes are not able to be captured in the fetch-limited wave estimate. Another source of difference comes from the pool of sample typhoons considered in the studies; the USACE (1981) considered 10 typhoon between 1961 and 1976; the SEI (2015) considered 30

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 33 typhoons and 23 severe tropical cyclones between 1950 and 2013; whereas GHD considered 88 typhoons between 1951 and 2019.

3.3.5 Effect of sea level rise

Effects of sea level rise were investigated by superimposing the projected 2070 SLR for intermediate emission, i.e. 2.3 ft., to the critical case, Case 4. Summary of results is shown in Table 13. Due to the additional water depth available from the SLR, Hs for wind waves and swell waves were increased by 0.3-0.5 ft. and 0.5-1 ft. respectively.

Table 13 Model results - Case 4 with 2070 SLR for intermediate emission

Point Water level (ft.

above MSL)

Current Speed (kt)

Wind wave Hs (ft.)

Wind wave Tp (sec)

Swell wave Hs (ft.)

Swell wave Tp (sec)

S5 9.5 0.8 4.8 3.6 3.0 12.8 S1 9.5 0.8 3.6 3.3 1.5 13.7 S2 9.5 0.8 3.5 3.2 1.3 13.6 S3 9.8 0.8 3.3 3.2 1.2 13.7 S4 9.8 0.8 3.3 3.3 1.2 13.7 O1 10.2 0.6 4.1 5.8 3.9 15.6 O2 10.2 0.2 3.9 3.5 3.1 15.6 O3 10.2 0.2 3.3 3.6 2.7 15.9 O4 10.2 0.2 3.0 3.6 2.4 15.7

3.4 Impacts from propeller wash

During ambient conditions, the project sites are impacted by two key influence factors:

Propeller wash, defined as a relatively high-velocity current in the shape of a cone generated by the spinning propellers in the water.

Boat wake, defined as a wave generated by the displacement of water as a boat travels through a body of water.

For the propeller wash, as the shoreline is orientated parallel to the direction of the travelling boat, the current generated by the propeller will be largely dissipated due to the viscosity of water as the velocity cone expands laterally. Once the cone diffuses and reaches the shoreline protection structure, its influence will be negligible compared to the governing design condition i.e. typhoon waves and the resulted forces, for the shoreline protection structure.

Similarly, boat wakes for small vessels are typically in the range of 0.6 ft. to 1.3 ft. with a period of approximately 2.2 seconds to 3.0 seconds (Edwards and Lord, 1998). Waves of this size are substantially smaller than the governing design condition and would not negatively affect the stability of the shoreline protection structure. This is demonstrated by the existing rock revetment just to the south of the remediated sheet pile at Smiling Cove, where the wakes generated by the passing boats have limited impact on the stability of the structure.

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 34

4. Conclusions and recommendations

4.1 Conclusions

The following are concluded from the simulated results of the wave study:

The dominant wind pattern in the Saipan region is the easterly trade wind, with seasonal trade winds from the northeast from November to March and easterly winds from May to October. The occurrence of wind speeds greater than 26 kts (44.5 ft/sec) are rare (~1%), and may be associated with typhoon events.

Extreme wind events at the project sites are generally associated with typhoons. Eighty-eight (88) historical typhoons were identified between the years 1951 and 2019 that had a minimum pressure of 950 hPa which passed the project sites within a 185 mi. (~300 km) radius. On average, one or two typhoons per year affecting the project sites occurred.

During ambient conditions, the majority of the offshore significant wave heights Hs are less than 10 ft. Most of the offshore waves have a peak wave period between 6 and 15 seconds and approach from the NNE to ENE directions.

The water levels at the offshore beyond the reef edge during typhoon events increased between 1.4 ft. and 3.1 ft., corresponding to a 2-yr ARI up to a 100-yr ARI event. This increased water level includes both tide and storm surge components.

Offshore typhoon waves are swell dominant, with periods ranging from 8 to 17 seconds.

The offshore typhoon significant wave heights range between 26.7 ft. and 59.8 ft.

corresponding to a 2-yr ARI up to a 100-yr ARI event.

Four water level and wave height combinations as described in Table 8 were investigated in this study. Case 4, which is the combination of 100-yr ARI wave height and 20-yr ARI water level, presents the worst-case scenario for Smiling and Outer Coves.

The wave climate inside the Tanapag Lagoon and at the project sites is largely influenced by the reef geometry including its width, foreshore slope, elevations of the reef crest, and surface roughness. Model results show that in the cases examined in this study, swell waves approaching from the offshore regions were mostly dissipated by the reef edge with its significant wave heights reduced by up to 91%; inside the Lagoon, wind-generated short period waves governed the wave climate. Due to wave set-up by the broken waves, the water levels inside the Lagoon increased by up to 300% (or 5 ft.).

In the absence of site-specific wave measurements, the GHD model results were compared to historical studies prepared by US Army Corps of Engineers and Sea Engineering Inc. for nearby locations. The comparison suggests the results were generally in agreement between the studies, with small discrepancies mainly contributed by the differences in water depths between the sites, the methodologies adopted by the studies, and the size of typhoon samples considered in the statistical analysis. For the purpose of designing a shoreline protection measure for a normal-importance infrastructure, GHD considers the model results acceptable in conjunction with the use of safety factor for the structural design.

The project sites are also exposed to hazards due to rising sea levels. As demonstrated by Case 4, the additional water depth provided from the SLR reduces wave breaking, thus allowing larger waves to penetrate to the project sites.

GHD | National Park Service - Repair of Shoreline Barriers at Smiling and Outer Coves PMIS#236143, 11214635 | 35

Boat wakes and propeller wash generated by a passing boat during ambient condition have an order of magnitude smaller impact…