Rates of subsidence and relative sea level rise in the Hawaii Islands

1 Introduction

For the study “Doubling of coastal erosion under rising sea level by mid-century in Hawaii” (Tiffany et al., 2015), a simple model is used to assess future erosion hazards under higher sea levels, taking into account historical changes of Hawaii shorelines and the projected acceleration of sea level rise predicted by the Intergovernmental Panel on Climate Change (IPCC). The results of the study by Tiffany et al. (2015) indicate that the coastal erosion of Hawaii’s beaches will double by mid-century because of global warming. As we have shown [18-44], as many others did, for example [1-5, 7, 8, 12-14, 50-52], the sea levels are slow rising and not accelerating worldwide and in the United States, and in the specific of the Hawaii Islands, as it will be shown, they have been actually decelerating over the last 3 decades. Therefore, the cause of the coastal erosion of the Hawaii Islands has to be searched elsewhere. Aim of the paper is to provide the latest estimations of relative sea level rise and subsidence rates in the Hawaii Islands.

2 Global sea level data from the psmsl global surveys

Global relative (sea level position vs. tide gauge instrument) sea level data is considered first. The latest 2015 global sea level survey by the Permanent Service on Mean Seal levels [46] tells us the sea levels are rising or falling in the surveyed 570 tide gauges of +1.04 ± 0.60 mm yr⁻¹ in the naïve average. The maximum rate of rise is + 10.25 mm yr⁻¹, the minimum −17.66 mm yr⁻¹, as the sea levels rise or fall very differently moving from one geographical area to the other and a global mean sea level rise is pure nonsense. The average record length is 60 years, with a maximum of 188 years and a minimum of 27 years. As the sea levels are characterized by inter-annual, decadal and multi-decadal oscillations about a longer term trend where the subsidence of the tide gauge is the largest component, it makes sense to consider only the 206 tide gauges of record length exceeding the 60 years. In this case, the naïve average become + 0.41 ± 0.29 mm yr⁻¹, while maximum and minimum become + 9.06 and −13.25 mm yr⁻¹ respectively. If we do consider only the 108 tide gauges of record length exceeding the 80 years, the naïve average become +0.19 ± 0.21 mm yr⁻¹, while maximum and minimum become +6.75 and −8.06 mm yr⁻¹ respectively. Comparison with previous surveys is not immediate, as the surveys do not include always the same tide gauges, and not always the rate of rise is computed in the same way. The changes of sea level rise in every tide gauge exceeding the 60 years is however negligible, and on average virtually and practically zero. Further details of the computation and the quality of the data set are provided on the PSMSL website.

3 Local sea level pattern from the noaa us surveys

More local relative sea level data is then considered. The latest 2015 United States sea level survey by the National Oceanic and Atmospheric Administration [15] tells us the sea levels are rising or falling in the surveyed 142 tide gauges of +1.74 ± 0.68 mm yr⁻¹ in the naïve average. The maximum rate of rise is +9.65 mm yr⁻¹, the minimum −17.59 mm yr⁻¹, as the sea levels rise or fall very differently moving from one geographical area to the other also in the United States and a United States mean sea level rise is similarly to the global mean sea level pure nonsense. The longest record is 158 years, the shortest 21 years the average record length 61 years. By only considering the 62 tide gauges of length exceeding the 60 years, the average sea level rise is +1.84 ± 0.29 mm yr⁻¹, while maximum and minimum are +9.03 and −17.59 mm yr⁻¹. By requiring more than 80 years, the 37 tide gauges have now a naïve average relative rate of rise of +2.16 ± 0.23 mm yr⁻¹, and maximum and minimum of +6.34 and −2.27 mm yr⁻¹ respectively. This is the result of more long term tide gauges covering the east coast of the US vs. the west coast of the US, the Hawaii Islands and Alaska for historical reasons. Comparison with previous surveys is not immediate, as the surveys do not include always the same tide gauges, and not always the rate of rise is computed in the same way. The changes of sea level rise in every tide gauge exceeding the 60 years is however negligible, and on average virtually and practically zero. Further details of the computation and the quality of the data set are provided on the NOAA website.

4 Analysis of the hawaiian tide gauges’ time series

Further localized relative sea level data is then considered, specific for the Hawaii Islands. Table 1 (data from [15]) and Fig. 1 (data from [48, 49]) presents the latest sea level rises computed by analysing the data recorded by the Hawaiian tide gauges. An indication of the possible subsidence of the tide gauges is also presented. This figure presents the monthly average mean sea level recorded by the Honolulu (PSMSL ID 155), Hilo, Hawaii Island (PSMSL ID 300) and Kahuli Harbour, Maui Island (PSMSL ID 521) tide gauges, the three longest tide gauges of the Hawaii Islands. Data are from PSMSL, 2015b.

Fig. 1

Mean sea level recorded by the tide gauges of Honolulu, Hilo, Hawaii Island and Kahului Harbor, Maui Island and local forecast to 2050. Recorded data are from PSMSL, 2015b. Images of inland GPS dome velocities are from SONEL, 2015. The subsidence of the GPS dome “HNLC” nearby the Honolulu tide gauge is −0.36 mm yr⁻¹. The subsidence of the GPS dome “HILO” nearby the Hilo tide gauge is −1.92 mm yr⁻¹. Despite the local subsidence of the tide gauge also evidenced by the oscillations of the monthly average mean sea level about a trend line of positive slope in not suspicious times, the latest short term trend over the period of the satellite monitoring is characterized by a mostly negative acceleration.

Despite the local subsidence of the tide gauge evidenced by the oscillations of the monthly average mean sea level about a trend line of positive slope in not suspicious times, +1.50, +3.58 and +2.12 mm yr⁻¹ respectively, the latest short term trend over the period of the satellite monitoring is strongly reduced in the three locations, +0.25, +1.17 and +0.43 mm yr⁻¹ respectively. This is however only the result of the multi-decadal oscillations, similarly to the East Coast “hot-spots”.

As shown in [18], the relative rates of rise of sea levels presently above the longer term trend along the east coast of the United States are the result of the phasing of multi-decadal oscillations. Similarly, [24], the relative rates of rise of sea levels presently below the longer term trend along the West Coast of the United States and Alaska are the result of the different phasing of multi-decadal oscillations.

If the time series of the monthly average mean sea levels de-trended vs. the linear trend are decomposed with a series of sinusoidal components, the amplitude, periodicity and phase of the sinusoids changes from area to area, with however often quasi-20 and quasi-60 years’ periodicities evidenced [20].

The images of the inland GPS dome velocities are from SONEL, 2015. The subsidence of the GPS dome “HNLC” nearby the Honolulu tide gauge is −0.36 mm yr⁻¹. The subsidence of the GPS dome “HILO” nearby the Hilo tide gauge is −1.92 mm yr⁻¹. A better description of the subsidence in the Hawaii Islands is provided in a following section by using the data from JPL, 2016 that monitors 20 GPS domes in the Hawaii Islands. For JPL, HNLC has a subsidence rate of −0.410 mm yr⁻¹ and HILO −1.557 mm yr⁻¹.

Despite the local subsidence of the tide gauge, that is usually larger than the subsidence of a nearby inland fixed point due to erosion/compaction, and it is possibly better evidenced by the oscillations of the monthly average mean sea level about a trend line of positive slope in not suspicious times, the latest short term trend over the period of the satellite monitoring is characterized by a mostly negative acceleration. Fig. 2 presents the mean sea level trends computed for the Hawaii islands with 60 and 30 years’ time windows. Images are from PSMSL [48].

Fig. 2

Mean Sea Level trends (in mm yr⁻¹) measured for the Hawaii islands with (top) 60 years’ time window 1954 to 2013, and (bottom) 30 years’ time window 1984–2013. Images are from PSMSL, 2015b. Moving north-westward in the top image, the trends reduce from the +2.54 mm yr⁻¹ of Hilo, Hawaii Island (PSMSL ID 300), to the +1.78 mm yr⁻¹ of Kahului Harbor, Maui Island (PSMSL ID 521), then to the +1.23 mm yr⁻¹ of Honolulu (PSMSL ID 155), the +1.19 mm yr⁻¹ of Mokuole Island (PSMSL ID 823) and the +1.40 mm yr⁻¹ of Nawilivili Bay, Kauai Island (PSMSL ID 756). In the shortest time window of the bottom image, the trends are +1.46 +2.54 mm yr⁻¹ in Hilo (-1.08 mm yr⁻¹), +1.24 mm yr⁻¹ in Kahului Harbor (-0.54 mm yr⁻¹), +1.02 mm yr⁻¹ in Honolulu (-0.24 mm yr⁻¹), +1.36 mm yr⁻¹ of Mokuole Island (+0.17) and +1.32 mm yr⁻¹ in Nawilivili Bay (-0.08 mm yr⁻¹). This result shows the significant increasing sinking of Hilo, located southeast.

Moving north-westward in the top image, 60 years’ time window, the trends reduce from the +2.54 mm yr⁻¹ of Hilo, Hawaii Island (PSMSL ID 300), to the +1.78 mm yr⁻¹ of Kahului Harbor, Maui Island (PSMSL ID 521), then to the +1.23 mm yr⁻¹ of Honolulu (PSMSL ID 155), the +1.19 mm yr⁻¹ of Mokuole Island (PSMSL ID 823) and the +1.40 mm yr⁻¹ of Nawilivili Bay, Kauai Island (PSMSL ID 756).

In the shortest time window of the bottom image, 30 years’ time window, the trends are +1.46 mm yr⁻¹ in Hilo (Δ =-1.08 mm yr⁻¹), +1.24 mm yr⁻¹ in Kahului Harbor (Δ =-0.54 mm yr⁻¹), +1.02 mm yr⁻¹ in Honolulu (Δ =-0.24 mm yr⁻¹), +1.36 mm yr⁻¹ of Mokuole Island (Δ =+0.17) and +1.32 mm yr⁻¹ in Nawilivili Bay (Δ =-0.08 mm yr⁻¹). This result shows the significant increasing sinking of Hilo, located south-east.

As the global relative sea level rise has zero acceleration, the most likely sea level pattern in the Hawaii until 2050 is therefore the local trend described so far with superimposed oscillations of up to quasi 60 years' periodicity forecasted to 2050, as shown in Fig. 1.

5 Discussion of atmospheric and sea level interconnections

Teleconnections are climate anomalies in atmospheric science being related to each other at large distances. Teleconnections were first noted through correlation between time series of atmospheric pressure, temperature and rainfall and are the building block of climate variability. While ocean and atmospheric patterns are certainly related, we do not certainly expect that the sea levels oscillate same of atmospheric pressure, temperature or rainfall. We already wrote [19, 20, 24], as many other did, that the time series from the Pacific tide gauges exhibit strong multi-decadal variabilities of quasi-20 and quasi-60 years, with however variables phases and amplitudes across the Pacific ocean. Relevant teleconnections for the Pacific are El Nińo/Southern Oscillation (ENSO), a periodic fluctuation in sea surface temperature and atmospheric pressure in the equatorial Pacific Ocean, and the Pacific Decadal Oscillation (PDO) a long-lived El Nino-like pattern of Pacific climate variability. The Southern Oscillation Index (SOI) is based on the observed sea level atmospheric pressure differences between Tahiti and Darwin and describe the state of the Southern Oscillation during El Nińo and La Nińa occurrences.

As shown in Fig. 3, there is not too much of correlation between the normalized de-trended monthly average mean sea levels in Honolulu (data from PSMSL [46]) and the normalized SOI (data from NOAA [16]). The correlation is certainly better in between the normalized de-trended monthly average mean sea levels in Honolulu and the normalized PDO (data also from NOAA [16]), even if the differences in between the 36 months moving averages are still substantial. If we consider the periodogram of the normalized de-trended monthly average mean sea levels in Honolulu and the normalized PDO, by using the Wessa software [53], the signature of the two time series is indeed very different, with the normalized PDO time series having significant periodicities about 20, 30, 40 and 80 years, but the normalized de-trended monthly average mean sea levels in Honolulu time series having only the about 20 years periodicity detectable. While the PDO is perhaps the most influential of the oscillations of the area and this oscillation certainly influences events in the Hawaii Islands, the PDO alone does not certainly explain the sea level pattern in the Hawaii Islands.

Fig. 3

Oscillation about the linear trend line of the mean sea level recorded by the Honolulu (PSMSL ID 155) tide gauge, plus time series of the Pacific Decadal Oscillation (PDO) and the Southern Oscillation Index (SOI) over the last 60 years. The three parameters are normalized vs. the difference in between their maximum and minimum values. Periodogram of the normalized oscillation about the linear trend line of the mean sea level recorded by the Honolulu tide gauge and the PDO.

The relative sea levels oscillations are much more complicated phenomena and cannot be explained by only considering the teleconnections of atmospheric patterns.

6 Discussion of relative vs. absolute sea levels

The paper uses “relative” sea level data. The tide gauge results are the most reliable and accurate information we do have to understand the complex oscillatory pattern of the sea levels. “Relative” means that the sea levels are measured relative to the tide gauge instrument that may be subject to subsidence or uplift. As the absolute vertical position of the tide gauge instrument is not known with accuracy or more often is not known at all, the concept of “absolute” sea levels, despite being popular, is therefore misleading.

The sea level rise attributed to global warming that is mentioned in many recent papers may be verified by the relative sea level results. If the monthly averaged relative mean sea level results are distributed since many decades about the same linear trend line, then the global warming contribution is shown to be negligible without any need of computing absolute sea level rises. Lacking a time rate of change of the relative sea level velocity at the tide gauge, i.e. a sea level acceleration, the claim global warming is contributing to the rate of rise of sea levels lack of any evidence.

If the sea levels are not rising that much because of melting of ices or thermal expansion is possibly because the actual warming occurring since the end of the last little ice age is actually quite mild, and the time scale to warm the oceans to the deepest layers is much larger than the time scale to warm up the surface layer or the atmosphere.

The evidence that the temperatures are not presently warming at an increased rate and the sea levels are consistently not sharply rising while positively accelerating may be found in [18-44, 50-52].

It is a fact that the satellite lower troposphere temperature (LTT) as properly monitored by satellites in the Remote Sensing Systems (RSS) and University of Alabama Huntsville (UAH) data sets is not warming since now 17 years, while the compilations of ground thermometer results are only rising by arbitrary continuous correction of the past temperatures made cooler or by biasing effects as urban heat island (UHI). This is perfectly consistent with the result from the world tide gauges, as surveyed by the Permanent Service on Mean Sea Levels (PSMSL) or the National Oceanic and Atmospheric Administration (NOAA), as the relatively stable sea level rises have no acceleration component over the same 17 years.

The stable sea level pattern is actually also supported by the satellite gravimeter or altimeter experiments returning slow rising not accelerating global sea levels before arbitrary correction by globally spreading the glacial isostatic adjustment (GIA). Also other indices, as the sea ice extension that is now turning globally stable as the Arctic sea ice shrinking is more than compensated by the Antarctic sea ice expansion, further support the fact that thermal expansion and ice melting have very little to contribute to sea level rises. All these results may be found summarized on a monthly basis in [9].

During this century, the sea levels have been rising in the Hawaii Island because of subsidence but at a reduced rate. There is therefore no reason to believe real the purely computational speculation of significant sea level rises by 2050 adopted in [49], as a likely scenario for the Hawaii Islands. The most likely sea level rise by 2050 in the Hawaiian Island is indeed a mere 89 millimeters vs. the values of the year 2000.

The major erosion problem for the Hawaii Islands is not because the anthropogenic carbon dioxide emission is changing the weather, but possibly because of the very well-known sinking of the volcanic islands. Therefore, te claim coastal erosion may double in the Hawaii by 2050 because of global warming lack of any evidence. If the rate of subsidence will not change, the costal erosion will increase in the next 34 years at about the same rate of the past 34 years.

7 Sinking of the land vs. rising of the sea

The sea level rise in the Hawaii Island due to thermal expansion and melting of ices on land does not seem dramatic. Therefore, the cause of the coastal erosion of the Hawaii Islands to the present time and to 2050 has to be searched elsewhere. As every geologist, or even every traveller visiting the island with minimal background information (for example Hawaii-guide.com [6]) knows, the Hawaiian Islands are situated near the middle of the Pacific Plate on top of a volcanic hot spot moving north-westward. This movement of the Pacific Plate over a local volcanic hot spot is what has produced the series of islands that is continually evolving. The Haleakala volcano in the island of Maui was very likely not only joined to the West Maui Mountains as it is today but also forming a single land mass that was combined with the today islands of Lanai, Molokai, and Kahoolawe, known as Maui Nui. The sinking of Maui Nui was the result of the volcanic body moved away from the Hawaiian hot spot translating in the lack of volcanic up-building combined with subsidence into the ocean floor. A similar fate awaits the Big Island in due time as the hot spot moves away. The geologic “circle-of-life” beyond the Hawaiian hot spot may be even more complicated, as for example recently shown in [17].

In addition to SONEL, also the National Aeronautics and Space Administration (NASA) Jet Propulsion Laboratory (JPL) proposes GPS time series (JPL [10]). The JPL GPS time series are quite extensive for the Hawaii Islands, totalling 20 stations.

The GPS is a constellation of 30 satellites used for navigation and precise geodetic position measurements. Data from receivers are analysed to produce time series got vertical velocity and horizontal velocities. The technique is rapidly evolving, and the inaccuracies of the estimations still substantial until recently are now dramatically reducing. Fig. 4 presents the horizontal velocities, mostly due to motion of the Earth’s tectonic plates, and the locations were the vertical velocities are also computed for the Hawaii Islands. The horizontal velocities are represented by lines extending from each site location.

Fig. 4

Horizontal GPS velocities for the Hawaii Islands (JPL, 2016). Image downloaded July 30, 2016. Moving south-east wards, there are 3 GPS stations in Kauai, one in Oahu and Maui, and finally 15 in the Island of Hawaii, 12 of them along the south-eastern coast.

Fig. 4 also presents the locations of the GPS stations. Fig. 5 presents the GPS time series of the horizontal and vertical velocities in these locations.

Fig. 5

GPS time series of horizontal and vertical GPS velocities for the Hawaii Islands (JPL, 2016). Images downloaded July 30, 2016. Moving south-east wards, there are 3 GPS stations in Kauai, one in Oahu and Maui, and finally 15 in the Island of Hawaii, 12 of them along the south-eastern coast.

In Kauai, KOK1 KOKB and LHUE, the horizontal velocities are about 34–35 mm yr⁻¹ southwards and 61–62 mm yr⁻¹ eastwards. Subsidence is −1.276 mm yr⁻¹ in KOK1 and −2.064 mm yr⁻¹ in LHUE. In KOKB the subsidence is greatly reduced, actually a minimal uplift. In Oahu, HNLC has similar horizontal velocities and a subsidence velocity of −0.410 mm yr⁻¹. In Mahui, MAUI has similar horizontal velocities and a subsidence velocity of −0.416 mm yr⁻¹ very close to HNLC. In the Island of Hawaii, the 3 stations in the north-center, UPO1, HILO and MKEA have similar horizontal velocities of about 34–35 mm yr⁻¹ southwards and 63 mm yr-1 eastwards. Subsidence is stronger in UPO1, rated at −2.937 mm yr⁻¹, still substantial in HILO, rated at −1.557 mm yr⁻¹, and finally negligible in MKEA, where it is actually mentioned a small uplift. Along the south-east coast of this Island, may stations are located nearby, but offering very different values, sometimes extreme, of subsidence or uplift, JOKA, NPIT, CHUD, NPOC, KAMO, JCUZ, OKIT, MMAU, PG2R, PGF5, PGF4, PGF6, to represent the instability of this area.

The result of Fig. 5 evidences the need of properly accounting for the motion of the tide gauge instrument to deal with “absolute” sea level rises, and the difficulties in achieving this goal by GPS monitoring, struggle further exacerbated by the lack of a precise surveying of the relative position tide gauge vs. GPS dome.

Certainly, the results of a global glacial isostatic adjustment (GIA) model are inadequate to represent the vertical velocities of the tide gauges, and we are still very far from achieving accurate global representations of land velocities through measurements that may only follow a further development of the GPS monitoring. The assumption that along the coast the land sinks as the sea level rises, or the land rises as the sea level reduces, is unfortunately a more accurate measurement of the vertical land velocity along the coast than the GIA simulations.

8 Conclusions

The world relative sea levels are marginally rising and without any significant acceleration component. In the Hawaii islands, the relative sea levels are rising at a reducing rate due to the phasing of the multi-decadal oscillations.

The GPS monitoring of horizontal and vertical land velocities is still far from being accurate, but it is certainly progressing. The historical relative sea level rates of rise and the vertical velocities of inland GPS domes suggest the relative sea level rise is primarily due to the sinking of the tide gauge instrument.

The “circle of life” of the Hawaii Islands moving north-westward volcanic hot spot with the embedded land subsidence is the cause of localized sea level rise and coastal erosion varying from one island to the other.

There is no evidence coastal erosion may double in the Hawaii by 2050 because of anthropogenic global warming. This is only the usual extravagant claim of the one-cause-fits-all approach relating every event to the anthropogenic carbon dioxide emission.