A daily to seasonal Arctic sea ice forecasting AI

2021 ◽  
Author(s):  
Tom R. Andersson ◽  
J. Scott Hosking ◽  
Eleanor Krige ◽  
Maria Pérez-Ortiz ◽  
Brooks Paige ◽  
...  
Keyword(s):  
Sea Ice ◽  
Initial Conditions ◽  
Arctic Sea Ice ◽  
Dynamical Model ◽  
The Arctic ◽  
Climate Simulation ◽  
Ice Thickness ◽  
Sea Ice Concentration ◽  
Sea Ice Thickness ◽  
Arctic Sea

<p>Arctic sea ice forecasting is a major scientific effort with fundamental challenges at play. To address such challenges, we have developed a physics-informed, data-driven sea ice forecasting system, IceNet, which outperformed a leading dynamical model (ECMWF SEAS5) in monthly-averaged forecasts of pan-Arctic sea ice concentration. IceNet adopted a U-Net deep learning architecture and was trained on over 2,000 years of CMIP6 climate simulation data. Despite its state-of-the-art seasonal forecasting skill at lead times of 2-6 months, IceNet has two main limitations. First, it could not outperform the dynamical model in short-range (1-month) forecasts. This is partly caused by IceNet operating on monthly-averages, which smears the initial conditions and weather phenomena that can dominate predictability at short time scales. Second, IceNet is afflicted by the ‘spring predictability barrier’ that affects all long range forecasts of summer. This predictability barrier arises primarily due to the importance of melt-season ice thickness conditions on summer sea ice. Here we present our early findings from IceNet2, which attempts to alleviate these issues by operating on daily-averages and including sea ice thickness as an input variable. IceNet2 paves the way for our efforts to aid the Arctic conservation community by developing the first public, operational sea ice forecasting AI.</p>

Sea ice and atmospheric potential predictability in coupled GCMs

2021 ◽  
Author(s):  
Daniela Flocco ◽  
Ed Hawkins ◽  
Leandro Ponsoni ◽  
François Massonnett ◽  
Daniel Feltham ◽  
...  
Keyword(s):  
Sea Ice ◽  
Initial Conditions ◽  
Complete Information ◽  
Arctic Sea Ice ◽  
Prediction Skill ◽  
The Arctic ◽  
Ice Thickness ◽  
Sea Ice Concentration ◽  
Potential Predictability ◽  
Sea Ice Thickness

<p>Assimilation of sea ice concentration satellite products has successfully been used to initialize sea ice models and coupled NWP systems. Sea-ice thickness observations, being much less mature, are typically not assimilated. However, many studies suggest that initialization of winter sea-ice thickness could lead to improved prediction of Arctic summer sea ice. We have examined the potential for sea ice thickness observations to improve forecast skill on timescales from days to a year ahead in two state-of-the-art coupled GCMs.</p><p>Here we examine the influence of Arctic sea-ice thickness observations on the potential predictability of the sea-ice and atmospheric circulation using idealised ‘data denial’ experiments. We perform paired sets of ensembles with the HadGEM3 and EC-Earth GCMs using different initial conditions retrieved from present-day control runs.</p><p>One set of ensembles start with complete information about the sea-ice conditions and is treated as “truth”, and one set has degraded sea ice information. We investigate how the pairs of ensembles, all started in January, predict the subsequent evolution of the sea-ice state, sea level pressure and circulation within the Arctic with the aim of quantifying the value of sea-ice observations for improving predictions.</p><p>We show that accurate initialization of sea ice thickness improves the model prediction skill during the first month of simulation and that several sea ice state and atmospheric variables present a re-emergence of skill in September. Prediction skill of several oceanic variables is also observed. The two models present a good agreement in terms of the regions where they show either a skill gain or loss.</p>

A data assimilation application for improving estimates of Arctic sea ice thickness variability and change since the turn of the 21st century

2021 ◽  
Author(s):  
Molly Wieringa ◽  
Cecilia Bitz
Keyword(s):  
Data Assimilation ◽  
Sea Ice ◽  
Initial Conditions ◽  
Arctic Sea Ice ◽  
Prognostic Models ◽  
Ice Thickness ◽  
Sea Ice Concentration ◽  
Sea Ice Thickness ◽  
Arctic Sea ◽  
Arctic Sea Ice Thickness

<p>Current sea ice prediction systems exhibit significant room for improvement compared to idealized estimates of sea ice predictability, a gap that could be closed by improving the initial conditions provided to prognostic models. Sea ice volume, the area-weighted integral of sea ice thickness (SIT), in particular, demonstrates long initial value predictability; in other words, accurate forecasting of Arctic sea ice requires highly accurate SIT initial conditions. Continuous records of SIT are, unfortunately, few and far between. To address this conundrum, we have explored applications of the Data Assimilation Research Testbed (DART) to constrain the Los Alamos Sea Ice Model (CICE5) within the Community Earth System Model using satellite-derived SIT observations from 2003 to present day. Our data assimilation system has been fine-tuned using new and highly accurate freeboard measurements from NASA’s ICESat-2 mission. Using SIT information alone, we generate two assimilation products: the first using DART with CICE5 and the second with an offline assimilation method. We compare these products to one another and to the community standard SIT record, PIOMAS. Future work will introduce multivariate assimilation of SIT with other sea ice variables, including sea ice concentration, sea ice skin temperature, and sea surface temperature.</p>

scholarly journals Near Real Time Arctic sea ice thickness and volume from CryoSat-2

2016 ◽  
Author(s):  
R. L. Tilling ◽  
A. Ridout ◽  
A. Shepherd
Keyword(s):  
Sea Ice ◽  
Real Time ◽  
Arctic Sea Ice ◽  
Arctic Climate ◽  
The Arctic ◽  
Ice Thickness ◽  
Satellite Orbits ◽  
Sea Ice Thickness ◽  
Arctic Sea ◽  
Arctic Sea Ice Thickness

Abstract. Timely observations of sea ice thickness help us to understand Arctic climate, and can support maritime activities in the Polar Regions. Although it is possible to calculate Arctic sea ice thickness using measurements acquired by CryoSat-2, the latency of the final release dataset is typically one month, due to the time required to determine precise satellite orbits. We use a new fast delivery CryoSat-2 dataset based on preliminary orbits to compute Arctic sea ice thickness in near real time (NRT), and analyse this data for one sea ice growth season from October 2014 to April 2015. We show that this NRT sea ice thickness product is of comparable accuracy to that produced using the final release CryoSat-2 data, with an average thickness difference of 5 cm, demonstrating that the satellite orbit is not a critical factor in determining sea ice freeboard. In addition, the CryoSat-2 fast delivery product also provides measurements of Arctic sea ice thickness within three days of acquisition by the satellite, and a measurement is delivered, on average, within 10, 7 and 6 km of each location in the Arctic every 2, 14 and 28 days respectively. The CryoSat-2 NRT sea ice thickness dataset provides an additional constraint for seasonal predictions of Arctic climate change, and will allow industries such as tourism and transport to navigate the polar oceans with safety and care.

Sea-ice freeboard or thickness? Design choices in the context of data assimilation in the coupled numerical prediction system EC-Earth3 for seasonal Arctic sea ice prediction

2021 ◽  
Author(s):  
Francois Massonnet ◽  
Sara Fleury ◽  
Florent Garnier ◽  
Ed Blockley ◽  
Pablo Ortega Montilla ◽  
...  
Keyword(s):  
Data Assimilation ◽  
Sea Ice ◽  
General Circulation ◽  
Seasonal Prediction ◽  
Arctic Sea Ice ◽  
Ice Thickness ◽  
Sea Ice Concentration ◽  
Sea Ice Thickness ◽  
Arctic Sea ◽  
Ice Concentration

<p>It is well established that winter and spring Arctic sea-ice thickness anomalies are a key source of predictability for late summer sea-ice concentration. While numerical general circulation models (GCMs) are increasingly used to perform seasonal predictions, they are not systematically taking advantage of the wealth of polar observations available. Data assimilation, the study of how to constrain GCMs to produce a physically consistent state given observations and their uncertainties, remains, therefore, an active area of research in the field of seasonal prediction. With the recent advent of satellite laser and radar altimetry, large-scale estimates of sea-ice thickness have become available for data assimilation in GCMs. However, the sea-ice thickness is never directly observed by altimeters, but rather deduced from the measured sea-ice freeboard (the height of the emerged part of the sea ice floe) based on several assumptions like the depth of snow on sea ice and its density, which are both often poorly estimated. Thus, observed sea-ice thickness estimates are potentially less reliable than sea-ice freeboard estimates. Here, using the EC-Earth3 coupled forecasting system and an ensemble Kalman filter, we perform a set of sensitivity tests to answer the following questions: (1) Does the assimilation of late spring observed sea-ice freeboard or thickness information yield more skilful predictions than no assimilation at all? (2) Should the sea-ice freeboard assimilation be preferred over sea-ice thickness assimilation? (3) Does the assimilation of observed sea-ice concentration provide further constraints on the prediction? We address these questions in the context of a realistic test case, the prediction of 2012 summer conditions, which led to the all-time record low in Arctic sea-ice extent. We finally formulate a set of recommendations for practitioners and future users of sea ice observations in the context of seasonal prediction.</p>

scholarly journals Estimating Arctic Sea Ice Thickness with CryoSat-2 Altimetry Data Using the Least Squares Adjustment Method

Sensors ◽  
2020 ◽  
Vol 20 (24) ◽  
pp. 7011
Author(s):  
Feng Xiao ◽  
Fei Li ◽  
Shengkai Zhang ◽  
Jiaxing Li ◽  
Tong Geng ◽  
...  
Keyword(s):  
Sea Ice ◽  
Least Squares ◽  
Arctic Sea Ice ◽  
The Arctic ◽  
Ice Thickness ◽  
Altimetry Data ◽  
Sea Ice Thickness ◽  
Arctic Sea ◽  
Least Squares Adjustment ◽  
Arctic Sea Ice Thickness

Satellite altimeters can be used to derive long-term and large-scale sea ice thickness changes. Sea ice thickness retrieval is based on measurements of freeboard, and the conversion of freeboard to thickness requires knowledge of the snow depth and snow, sea ice, and sea water densities. However, these parameters are difficult to be observed concurrently with altimeter measurements. The uncertainties in these parameters inevitably cause uncertainties in sea ice thickness estimations. This paper introduces a new method based on least squares adjustment (LSA) to estimate Arctic sea ice thickness with CryoSat-2 measurements. A model between the sea ice freeboard and thickness is established within a 5 km × 5 km grid, and the model coefficients and sea ice thickness are calculated using the LSA method. Based on the newly developed method, we are able to derive estimates of the Arctic sea ice thickness for 2010 through 2019 using CryoSat-2 altimetry data. Spatial and temporal variations of the Arctic sea ice thickness are analyzed, and comparisons between sea ice thickness estimates using the LSA method and three CryoSat-2 sea ice thickness products (Alfred Wegener Institute (AWI), Centre for Polar Observation and Modelling (CPOM), and NASA Goddard Space Flight Centre (GSFC)) are performed for the 2018–2019 Arctic sea ice growth season. The overall differences of sea ice thickness estimated in this study between AWI, CPOM, and GSFC are 0.025 ± 0.640 m, 0.143 ± 0.640 m, and −0.274 ± 0.628 m, respectively. Large differences between the LSA and three products tend to appear in areas covered with thin ice due to the limited accuracy of CryoSat-2 over thin ice. Spatiotemporally coincident Operation IceBridge (OIB) thickness values are also used for validation. Good agreement with a difference of 0.065 ± 0.187 m is found between our estimates and the OIB results.

scholarly journals Multidecadal Arctic sea ice thickness and volume derived from ice age

2020 ◽  
Vol 14 (4) ◽  
pp. 1325-1345 ◽  
Author(s):  
Yinghui Liu ◽  
Jeffrey R. Key ◽  
Xuanji Wang ◽  
Mark Tschudi
Keyword(s):  
Sea Ice ◽  
Global Climate ◽  
Arctic Sea Ice ◽  
Ice Age ◽  
The Arctic ◽  
Ice Thickness ◽  
Sea Ice Thickness ◽  
Ice Volume ◽  
Arctic Sea ◽  
Arctic Sea Ice Thickness

Abstract. Sea ice is a key component of the Arctic climate system, and has impacts on global climate. Ice concentration, thickness, and volume are among the most important Arctic sea ice parameters. This study presents a new record of Arctic sea ice thickness and volume from 1984 to 2018 based on an existing satellite-derived ice age product. The relationship between ice age and ice thickness is first established for every month based on collocated ice age and ice thickness from submarine sonar data (1984–2000) and ICESat (2003–2008) and an empirical ice growth model. Based on this relationship, ice thickness is derived for the entire time period from the weekly ice age product, and the Arctic monthly sea ice volume is then calculated. The ice-age-based thickness and volume show good agreement in terms of bias and root-mean-square error with submarine, ICESat, and CryoSat-2 ice thickness, as well as ICESat and CryoSat-2 ice volume, in February–March and October–November. More detailed comparisons with independent data from Envisat for 2003 to 2010 and CryoSat-2 from CPOM, AWI, and NASA GSFC (Goddard Space Flight Center) for 2011 to 2018 show low bias in ice-age-based thickness. The ratios of the ice volume uncertainties to the mean range from 21 % to 29 %. Analysis of the derived data shows that the ice-age-based sea ice volume exhibits a decreasing trend of −411 km3 yr−1 from 1984 to 2018, stronger than the trends from other datasets. Of the factors affecting the sea ice volume trends, changes in sea ice thickness contribute more than changes in sea ice area, with a contribution of at least 80 % from changes in sea ice thickness from November to May and nearly 50 % in August and September, while less than 30 % is from changes in sea ice area in all months.

scholarly journals On the timescales and length scales of the Arctic sea ice thickness anomalies: a study based on 14 reanalyses

2019 ◽  
Vol 13 (2) ◽  
pp. 521-543 ◽  
Author(s):  
Leandro Ponsoni ◽  
François Massonnet ◽  
Thierry Fichefet ◽  
Matthieu Chevallier ◽  
David Docquier
Keyword(s):  
Data Assimilation ◽  
Sea Ice ◽  
Arctic Sea Ice ◽  
Ice Thickness ◽  
Sea Ice Concentration ◽  
Length Scales ◽  
Sea Ice Thickness ◽  
Arctic Sea ◽  
Ice Concentration ◽  
Arctic Sea Ice Thickness

Abstract. The ocean–sea ice reanalyses are one of the main sources of Arctic sea ice thickness data both in terms of spatial and temporal resolution, since observations are still sparse in time and space. In this work, we first aim at comparing how the sea ice thickness from an ensemble of 14 reanalyses compares with different sources of observations, such as moored upward-looking sonars, submarines, airbornes, satellites, and ice boreholes. Second, based on the same reanalyses, we intend to characterize the timescales (persistence) and length scales of sea ice thickness anomalies. We investigate whether data assimilation of sea ice concentration by the reanalyses impacts the realism of sea ice thickness as well as its respective timescales and length scales. The results suggest that reanalyses with sea ice data assimilation do not necessarily perform better in terms of sea ice thickness compared with the reanalyses which do not assimilate sea ice concentration. However, data assimilation has a clear impact on the timescales and length scales: reanalyses built with sea ice data assimilation present shorter timescales and length scales. The mean timescales and length scales for reanalyses with data assimilation vary from 2.5 to 5.0 months and 337.0 to 732.5 km, respectively, while reanalyses with no data assimilation are characterized by values from 4.9 to 7.8 months and 846.7 to 935.7 km, respectively.

scholarly journals Presentation and evaluation of the Arctic sea ice forecasting system neXtSIM-F

2019 ◽  
Author(s):  
Timothy Williams ◽  
Anton Korosov ◽  
Pierre Rampal ◽  
Einar Ólason
Keyword(s):  
Sea Ice ◽  
Initial Conditions ◽  
Arctic Sea Ice ◽  
The Arctic ◽  
Forecast System ◽  
Sea Ice Concentration ◽  
Sea Ice Thickness ◽  
Daily Data ◽  
Ice Concentration ◽  
Forecasting System

Abstract. The neXtSIM-F forecast system consists of a stand-alone sea ice model, neXtSIM, forced by the TOPAZ ocean forecast and the ECMWF atmospheric forecast, combined with daily data assimilation. It was tested for the northern winter of 2018–2019 with different data being assimilated and was found to perform well. Despite drift not being assimilated in our system, we obtain quite good agreement between observations, comparing well to more sophisticated coupled ice-ocean forecast systems. The RMSE in drift speed is around 3 km/day for the first three days, climbing to about 4 km/day for the next day or two; computing the RMSE in the total drift adds about 1 km/day to the error in speed. The drift bias remains close to zero over the whole period from Nov 2018–Apr 2019. The neXtSIM-F forecast system assimilates OSISAF sea ice concentration products (both SSMI and AMSR2) and SMOS sea ice thickness by modifying the initial conditions daily and adding a compensating heat flux to prevent removed ice growing back too quickly. This greatly improved the agreement of these quantities with observations for the first 3–4 days of the forecast.

scholarly journals Suppression of Arctic sea ice growth by winter clouds and snowfall

2021 ◽  
Author(s):  
Won-il Lim ◽  
Hyo-Seok Park ◽  
Andrew Stewart ◽  
Kyong-Hwan Seo
Keyword(s):  
Sea Ice ◽  
Longwave Radiation ◽  
Arctic Sea Ice ◽  
The Arctic ◽  
Ice Thickness ◽  
Sea Ice Thickness ◽  
Ice Growth ◽  
Downward Longwave Radiation ◽  
Arctic Sea ◽  
The Impact

Abstract The ongoing Arctic warming has been pronounced in winter and has been associated with an increase in downward longwave radiation. While previous studies have demonstrated that poleward moisture flux into the Arctic strengthens downward longwave radiation, less attention has been given to the impact of the accompanying increase in snowfall. Here, utilizing state-of-the art sea ice models, we show that typical winter snowfall anomalies of 1.0 cm, accompanied by positive downward longwave radiation anomalies of ~5 W m-2 can decrease sea ice thickness by around 5 cm in the following spring over the Eurasian Seas. This basin-wide ice thinning is followed by a shrinking of summer ice extent in extreme cases. In the winter of 2016–17, anomalously strong warm/moist air transport combined with ~2.5 cm increase in snowfall decreased spring ice thickness by ~10 cm and decreased the following summer sea ice extent by 5–30%. Projected future reductions in the thickness of Arctic sea ice and snow will amplify the impact of anomalous winter snowfall events on winter sea ice growth and seasonal sea ice thickness.

Sign in / Sign up

Export Citation Format

Share Document