Storms and waves MCCIP Science Review 2020 132–157 132 Impacts of climate change on storms and waves relevant to the coastal and marine environment around the UK J. Wolf, 1 D. Woolf 2 and L. Bricheno 3 1 Physical Oceanographer, National Oceanography Centre, 6 Brownlow St, Liverpool, L3 5DA, UK 2 Reader in Marine Physics, Heriot-Watt University, Orkney Campus, The Old Academy, Back Road, Stromness, Orkney, KW16 3AW, UK 3 Marine Modeller, National Oceanography Centre, 6 Brownlow St, Liverpool, L3 5DA, UK EXECUTIVE SUMMARY We have updated the review by Woolf and Wolf (2013) by summarising the results of the IPCC AR5 report for storms and waves and then including more-recent work published since 2013. There are similar conclusions: wave- model results are controlled largely by the quality of the wind data used to drive them, and the forcing climate models have slightly improved in accuracy as well as resolution. In general, trends are obscured by wide natural variability and a low signal-to-noise ratio. Assessment of changes in storminess and waves over the last 200 years are limited by lack of data, while future projections are limited by the accuracy of climate models. Recent work has led to more insight in some areas. There are now more climate- and wave- model ensembles, more in-depth assessments of the results of CMIP5, and the CMIP6 project and IPCC AR6 assessments have started. There is a move towards higher-resolution models, which give better accuracy for simulation of tropical and extra-tropical storms. Further work is being done with coupled atmosphere-ocean-wave models, which give insight into key dynamic processes. There is evidence for an increase in North Atlantic storms at the end of the 20 th Century. Some projections for North Atlantic storms over the 21 st Century show an overall reduced frequency of storms and some indication of a poleward shift in the tracks, in the northern hemisphere (NH) winter, but there is substantial uncertainty in projecting changes in NH storm tracks, especially in the North Atlantic. Projections for waves in the North Atlantic show a reduction in mean wave height, but an increase in the most-severe wave heights. There is a likelihood of larger wave heights to the north of the UK as the Arctic sea ice retreats and leads to increased fetch. 1. INTRODUCTION Surface wind waves and storm-force winds can cause much damage in UK coastal waters, particularly in autumn and winter. Understanding the Citation: Wolf, J., Woolf, D. and Bricheno, L. (2020) Impacts of climate change on storms and waves relevant to the coastal and marine environment around the UK. MCCIP Science Review 2020, 132–157. doi: 10.14465/2020.arc07.saw Submitted: 09 2018 Published online: 15 th January 2020.
26
Embed
Impacts of climate change on storms and waves …Storms and waves MCCIP Science Review 2020 132–157 132 Impacts of climate change on storms and waves relevant to the coastal and
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Storms and waves
MCCIP Science Review 2020 132–157
132
Impacts of climate change on storms and
waves relevant to the coastal and marine
environment around the UK
J. Wolf,1 D. Woolf 2 and L. Bricheno 3
1 Physical Oceanographer, National Oceanography Centre, 6 Brownlow St, Liverpool, L3 5DA,
UK
2 Reader in Marine Physics, Heriot-Watt University, Orkney Campus, The Old Academy, Back
Road, Stromness, Orkney, KW16 3AW, UK 3 Marine Modeller, National Oceanography Centre, 6 Brownlow St, Liverpool, L3 5DA, UK
EXECUTIVE SUMMARY
We have updated the review by Woolf and Wolf (2013) by summarising the
results of the IPCC AR5 report for storms and waves and then including
more-recent work published since 2013. There are similar conclusions: wave-
model results are controlled largely by the quality of the wind data used to
drive them, and the forcing climate models have slightly improved in
accuracy as well as resolution. In general, trends are obscured by wide natural
variability and a low signal-to-noise ratio. Assessment of changes in
storminess and waves over the last 200 years are limited by lack of data, while
future projections are limited by the accuracy of climate models.
Recent work has led to more insight in some areas. There are now more
climate- and wave- model ensembles, more in-depth assessments of the
results of CMIP5, and the CMIP6 project and IPCC AR6 assessments have
started. There is a move towards higher-resolution models, which give better
accuracy for simulation of tropical and extra-tropical storms. Further work is
being done with coupled atmosphere-ocean-wave models, which give insight
into key dynamic processes.
There is evidence for an increase in North Atlantic storms at the end of the
20th Century. Some projections for North Atlantic storms over the 21st
Century show an overall reduced frequency of storms and some indication of
a poleward shift in the tracks, in the northern hemisphere (NH) winter, but
there is substantial uncertainty in projecting changes in NH storm tracks,
especially in the North Atlantic. Projections for waves in the North Atlantic
show a reduction in mean wave height, but an increase in the most-severe
wave heights. There is a likelihood of larger wave heights to the north of the
UK as the Arctic sea ice retreats and leads to increased fetch.
1. INTRODUCTION
Surface wind waves and storm-force winds can cause much damage in UK
coastal waters, particularly in autumn and winter. Understanding the
Citation: Wolf, J., Woolf, D.
and Bricheno, L. (2020)
Impacts of climate change on
storms and waves relevant to
the coastal and marine
environment around the UK.
MCCIP Science Review 2020,
132–157.
doi: 10.14465/2020.arc07.saw
Submitted: 09 2018
Published online: 15th January
2020.
Storms and waves
MCCIP Science Review 2020 132–157
133
characteristics of the mean and extreme wave climate, its variability, and
historical and projected future change is an important consideration for
sustainable development of coastal and offshore infrastructure, and
management of coastal resources and ecosystems. The effects of waves are
also critical to shipping; storm waves need to be avoided on shipping routes.
The reduction in summer sea-ice due to global warming is opening up the
Arctic sea routes to ships, but also increasing the fetch of waves in these
regions (Aksenov et al., 2017).
Except for tsunamis, waves are driven by the wind, with a nonlinear
relationship to wind-speed, fetch and duration over which the wind blows.
The largest waves in UK waters tend to be found on the Atlantic-facing coasts
where waves can be generated over large fetches in the ocean, and during the
period October to March (autumn and winter) when strong winds are more
intense and persistent. Many factors affect the height of waves in UK waters,
but for the Atlantic margin the persistence and strength of westerly winds are
particularly important, as well as the intensity and frequency of storms
(‘storminess’). In the North Sea, westerlies have a more-limited fetch, but
can still generate high waves. Northerly winds can generate high waves
particularly in the central and southern North Sea, whereas strong southerly
winds can generate high waves in the northern North Sea.
For the UK, the behaviour of the North Atlantic storm track is critical to
understanding storms and extreme waves. Decadal variability in terms of
storms and waves within the north-east Atlantic Ocean is mainly related to
the North Atlantic Oscillation (NAO), and affects the west-facing coasts of
the UK, but its effects can also be detected in the North Sea. The NAO index
is related to the pressure difference between the Azores and Iceland, which
influences the North Atlantic jet stream, storm tracks and blocking and
thereby affects winter wave climate over the North Atlantic (IPCC, 2013). A
positive NAO is usually accompanied by increased mean wave heights and
storminess in the Atlantic Margin and North Sea, whereas a negative NAO
tends to have the opposite effect (N.B. the NAO can also affect summer
weather, see Folland et al., 2009).
Significant Wave Height (SWH, often referred to using the variable HS)
represents a measure of the energy in the wave field, consisting of both wind–
sea and swell, and is approximately equal to the highest one-third of wave
heights. Other important parameters are wave period and wave direction,
which affect how waves impact the coast. Figure 1 shows an estimate of the
50-year return period SWH from Bricheno et al. (2015) to illustrate the
differences in wave exposure around the UK. It can be seen that the largest
waves are found in the north-west Approaches, north-west Scotland and the
Outer Hebrides. Lowest waves are seen in the more sheltered waters of the
eastern Irish Sea, southern North Sea and the eastern English Channel,
although wave height is not the only cause of danger. Short, steep seas of
Storms and waves
MCCIP Science Review 2020 132–157
134
lower height can be hazardous to small craft in storm conditions, even in
relatively short-fetch conditions.
Figure 1: 50-year return-period wave-height around UK from 10-year hind-cast
1999−2008 (from Bricheno et al., 2015). This figure is just an example to show the spatial
distribution of wave height around the UK but should not be referred to as the best estimate
of the 50-year return period as it has been extracted from too short a sample of model data.
In coastal waters, waves are affected by tidal currents and water depth, and
locally by coastal geometry and man-made structures. Coastal defences, such
as harbours, breakwaters and seawalls, are designed to dissipate wave energy
before it impacts the coast, as well as protecting against extreme water levels
caused by sea-level rise, tides and surges. Waves themselves can contribute
to raising the water level in a storm by means of wave setup, run-up and
overtopping (Prime et al., 2016). Waves will have different impacts on sandy
beaches, compared with rocky coasts, estuaries or saltmarshes. Some
background on coastal wave processes, monitoring and modelling is given in
Wolf (2016). Waves decrease in height as they shoal, due to energy
dissipation by bottom friction and wave breaking; this reduction in energy at
a particular site may diminish if sea level rises, unless the coastal morphology,
in areas of mobile sediment, can adapt at a similar rate. An important factor
with respect to coastal wave impact is ‘coastal squeeze’, in which the
nearshore depth profile is steepening as coastal defences are hardened on the
inland side and offshore water levels increase. Changes in this coastal zone
may be exacerbated by offshore aggregate extraction (although this is
regulated in the UK) or other man-made changes. In some areas there is now
a move towards the introduction of soft defences, such as beach recharge and
Storms and waves
MCCIP Science Review 2020 132–157
135
nature-based solution such as re-introducing saltmarshes (‘managed re-
alignment’).
Waves and storms are a significant feature of the global climate and have
been included in many assessments of climate including the latest assessment
(the Fifth Assessment Report) of the Intergovernmental Panel on Climate
Change (IPCC, 2013, hereafter referred to as ‘AR5’), which was published
since our last review, and consolidates the state of knowledge up to 2013. We
summarise the results of AR5 and discuss work carried out since then.
Here we focus on UK waters, but recognise that local changes in waves
depend on changes at much larger scales, since waves integrate wind energy
across ocean basins. In turn, large-scale patterns in winds are related to global
teleconnections that may manifest as inter-annual and decadal variability over
a regional scale, such as the North Atlantic Ocean. For the UK and Europe,
we are mainly concerned with extra-tropical cyclones (ETCs), also known as
‘mid-latitude storms’). However, we include a discussion of potential changes
in Tropical Cyclones (TCs) – termed ‘hurricanes’ in the North Atlantic,
because some TCs undergo ‘extratropical transition’ and can then track across
the North Atlantic to Europe and the UK. Note also that hurricane-force winds
(Beaufort scale Force 12 and above) are those with wind-speeds >32.6 m s-1,
which may also occur in events which are not actually hurricanes.
In general, we include only references published since the previous review in
2013, and not including those given in AR5, except where a topic was not
previously included. New topics include the use of coupled atmosphere-
ocean-wave models in the climate system and the emerging issue of
attribution of extreme events to climate change. We also extend the discussion
of storm and wave impacts at the coast and coastal adaptation to climate
change.
In Section 2 we mainly rely upon historical data, model hind-casts and climate
model reanalyses to understand what is already happening. In Section 3,
looking to the future, we rely on model projections. Confidence in historical
trends in storms and waves is generally low due to limited observations of
extreme events, and changes in observing methods. Future projections also
are subject to low confidence due to the dominance of natural variability in
the storm and wave climate.
2. WHAT IS ALREADY HAPPENING?
To understand past changes and trends in wave climate we need a long
time−series of observations, and where these are not available we may use
proxies, such as sediment deposits in peat bogs, to identify the occurrence of
past storms over palaeo timescales, e.g. Orme et al. (2017). Where there are
limited data available, as in the relatively recent past (since ~1800), we can
Storms and waves
MCCIP Science Review 2020 132–157
136
use model hind-casts e.g. WASA-Group (1998), STOWASUS-Group (2001),
NESS, NEXT and NEXTRA (Williams, 2005; 2008) and, increasingly over
the last decade, re-analyses combining models and observations. Re-analyses
use data assimilation in a dynamical model of the atmosphere and ocean,
which ideally maximises the benefit of the limited data, especially in the
earlier time periods, as well as providing dynamically consistent wind and
wave fields, allowing the calculation of wind and wave statistics in areas
where there are no data. New re-analyses have been released following
improvements in the models and/or data assimilation schemes from
operational Numerical Weather Prediction (NWP) centres. The re-analyses
differ in terms of the models and data assimilation methods used to produce
them, so they produce different results. However, some issues have been
found with inhomogeneities in long reanalyses, usually related to step
changes where new data assimilation is introduced, e.g. wave data from
altimeters in 1991 in ERA-Interim (Aarnes et al., 2015). The changing mix
of observations, and biases in observations and models, can introduce
spurious variability and trends into re-analysis output.
Since AR5 there have been many further studies, which are mentioned in
more detail where relevant in the following sections. The next IPCC
Assessment Report (AR6) has commenced. Waves are increasingly being
recognised as having an important role in air−sea fluxes and mixing processes
in the ocean as well as contributing to changes in mean water level (e.g.
Staneva et al., 2017). The use of coupled wave−atmosphere−ocean models is
increasing, although wave models have not yet been included in the Coupled
Model Intercomparison Project (CMIP), now in its 6th phase (CMIP6, Eyring
et al., 2016). The physics of atmospheric models is being improved
continually, with clouds, aerosols, atmospheric chemistry, biogeochemical
cycles and interactions with the ocean and cryosphere receiving attention,
some of which may have implications for storm initiation and evolution. For
example, Tamarin-Brodsky and Kaspi (2017) show that increased latitudinal
propagation in a warmer climate is due to stronger upper-level winds and
increased atmospheric water vapour. Stopa et al. (2016) discuss the
importance of waves in the marginal ice zone (MIZ) and Ardhuin et al. (2018)
examine the physics of interactions between waves and sea ice.
Some excerpts of AR5 are summarised in the next paragraphs for TCs, ETCs
and waves in the North Atlantic (details of spatial variation around the UK
are discussed elsewhere and note that IPCC definitions of likelihood and
confidence are adopted):
• Some high-resolution atmospheric models have realistically simulated
tracks and counts of TCs and models generally are able to capture the
general characteristics of storm tracks and ETCs with evidence of
improvement since the AR4.
Storms and waves
MCCIP Science Review 2020 132–157
137
• Storm track biases in the North Atlantic have improved slightly, but
models still produce a storm track that is too zonal and underestimate
cyclone intensity (Zappa et al., 2013a, b).
• There is low confidence in long-term (centennial) historical changes in
TC activity, after accounting for past changes in observing capabilities,
but over the satellite era (since the late 1980s), increases in the frequency
and intensity of the strongest storms in the North Atlantic are robust (very
high confidence). The cause of this increase is debated and there is low
confidence in attribution of changes in TC activity to human influence.
This is due to insufficient observational evidence, lack of physical
understanding of the links between anthropogenic drivers of climate and
TC activity and the low level of agreement between studies about the
relative importance of internal variability, and anthropogenic and natural
forcings (see AR5 sections 2.6.3, 10.6.1, 14.6.1).
• Over periods of a century or more, evidence suggests a slight decrease in
the frequency of TCs making landfall in the North Atlantic (in North
America, not Europe), once uncertainties in observing methods have been
considered. For ETCs, a poleward shift is evident in both hemispheres
over the past 50 years, with further, limited, evidence of a decrease in
wind storm frequency at mid-latitudes. Several studies suggest an
increase in intensity, but data sampling issues hamper these assessments.
• Global and regional time series of wind-wave characteristics are available
from buoy data, Voluntary Observing Ship (VOS) reports, satellite
measurements and model wave hind-casts. There is very strong evidence
that storm activity has increased in the North Atlantic since the 1970s.
• Positive regional trends in extreme wave heights have been reported at
several buoy locations since the late 1970s. Satellite altimeter
observations provide a further data source for wave height variability
since the mid-1980s. Model hind-casts based on 20CRv2 (spanning
1871–2010) and ERA40 (spanning 1958–2001) show increases in annual
and winter mean SWH in the North-East Atlantic, although the trend
magnitudes depend on the re-analysis products used (e.g. Stopa and
Cheung, 2014). Analysis of VOS observations for 1958–2002 reveals
increases in winter mean SWH over much of the North Atlantic, north of
45°N, with typical trends of up to 20 cm per decade.
19th−21st Century record − observations
Wave data have only been routinely collected by calibrated instruments, such
as wave buoys, since about 1950. Meteorological data collection has a longer
history and Sea Level Pressure (SLP) has been observed since the 19th
century, allowing construction of isobaric charts and analysis of winds and
storms from these data. Voluntary Observing Ships (VOS) have provided
some useful data on wind and waves since 1856 (Gulev et al., 2003; Gulev
and Grigorieva, 2004). Centennial time series of visually observed wave
height were derived from the International Comprehensive Ocean-
Atmosphere Data Set (ICOADS) along the major ship routes worldwide. In
Storms and waves
MCCIP Science Review 2020 132–157
138
the North Atlantic, and other basins, significant upward changes (up to 14
cm/decade) are observed, but only for the last 50 years and not for centennial
records. Long-term changes in wind wave height are closely associated with
the North Atlantic Oscillation (NAO) in the Atlantic. The reliability of such
data has been examined by Gulev et al. (2003).
In Woolf and Wolf (2010), we reviewed the observational data over the last
60 years, since reanalysis products at that time generally extended over that
era, and marine data greatly improved at that time, due to the advent of Ocean
Weather Stations (OWS) and other reliable sources of wind and waves data.
The measurement network has evolved in the last 70 years and particularly in
the last 30 years, since the advent of satellite wind and wave observations. In
the last update (Woolf and Wolf, 2013), we reviewed the original information,
plus longer time−series based on sea-level pressure. Here we add the
information gathered from VOS and more-recent, high-resolution, long re-
analysis datasets, which can maximise the benefits of earlier data, as well as
identifying biases introduced by changes in the methodology.
Existing wind and wave data sources around the UK can be found via the
MEDIN (Marine Environmental Data & Information Network) wave
metadata tool https://portal.medin.org.uk/portal/start.php, among others,
which allows discovery of wave and other marine data. The data sources
include wave buoys of the Wavenet monitoring network
https://www.cefas.co.uk/cefas-data-hub/wavenet/, operated by Cefas, the
Irish Marine Institute, the Met Office and the Channel Coastal Observatory
(CCO), originally focussed in the southern UK, but which also provides links
to other regions, namely the north-east, north-west, Anglia and the East
Riding of Yorkshire. In recent years, projects such as the EU-funded
COASTALT project (2009−2011), http://www.coastalt.eu/ has aimed to
recover more altimeter data in the nearshore zone, including waves.
A large amount of metocean data (including that for wind and waves) are
collected in situ, by, or for, major oil and gas companies, at considerable
cost. These companies have many offshore oil and gas fields scattered
worldwide in seas and on continental shelves, often in remote areas.
Metocean analyses provide them with essential information needed to
complement their working practices, such as in the design and engineering of
offshore installations and for the forecasting of meteorological events. The
System of Industry Metocean data for the Offshore and Research
Communities (SIMORC) is one source of long-term data