acquiring data down further. Such large geoscience datasets, coupled with latest supercomputer simulation outputs from global climate model intercomparison projects is now available for machine learning applications (Kay et al., 2015; Schneider et al., 2017; Reichstein et al., 2019).
In another related study, (O’Gorman & Dwyer, 2018) used random forests to parameterize moist convection processes to successfully emulate physical processes from expensive climate model outputs. (Scher, 2018) was also able to approximate the dynamics of a simple climate model faithfully after being presented enough data with deep learning.
Recently machine learning has demonstrated promise in resolving the largest source of uncertainty (Sherwood et al., 2014; IPCC, 2018) in climate projections, cloud convection (Rasp et al., 2018; Gentine et al., 2018). (Rasp et al., 2018) and (Gentine et al., 2018) demonstrated the use of deep learning in emulating sub-grid processes to resolve model clouds within simplified climate models at a fraction of the computational cost of high resolution physics models. These developments provide machine learning researchers opportunities to build models about ice flow and dynamics from satellite data, or using new video prediction techniques (Mathieu et al., 2016; Denton & Fergus, 2018) to predict changes in glaciers and ice dynamics.
In this work, our main contributions are (1) development of an unsupervised learning model to track ice sheet and glacier dynamics; and (2) introducing IceNet, a dataset that we make available for the community to serve as a first step in bridging gaps between machine learning and cryosphere climate research.
In this paper, we investigate on seven bands ranging from 0.43 m to 2.29
m(visible, near-infrared and shortwave light) with a resolution of 30 meters. The details of LANDSAT 8 can be found at https://www.usgs.gov/landresources/nli/landsat. In our dataset, we focus on a particular area at Antarctica with a latitude of 80
01’25” South and longitude 153
11’10” East, where the ice flow’s moving pattern is dominated by the Byrd Glacier (path 54, row 118 using worldwide reference system-2). The picture is shown in Figure 1.
Figure 1. (a) The region our dataset investigates. (b) Coastal signal(Band 1, 0.43m) collected by the LANDSAT 8 at 2015 November 22. The four corners contain no information.
Our dataset contains the satellite images ranging from November 2015 to February 2017, with total 10675 images and every image has 12 frames with the shape of 128 by 128; the interval between each frame ranges from two weeks to 9 month gaps, each pixel stands for a 30 meters by 30 meters region.
2.1. Labels
The images are denoted as where i is from 1 to 12 and the frames(subscenes) in each image are
, where
. For finding the next subscene, or chip, that matches the
best, we compare the
to a range of possible regions by calculating the correlation between two chips, the equation writes as:
where r and s are the two images and is the mean value. The ice flow is not static, moving areas of the large ice sheets
Figure 2. A larger subscene is selected in case of the previous subscene moving outside the original grid.
remain a challenge for tracing the ice flow. For tackling surface feature movement, we select a larger area by a scale factor c(c > 1) that centres around the previous subscene in case the pattern moving outside the previous grid, the most correlated one is chosen as the next subscene(the ground truth). The pipeline is shown in Figure 2.
We use a stochastic video generation with prior for prediction. The prior network observes frames xand output
and
of a normal distribution and is trained with by maxing:
Where and
are generated from convolutional LSTM.
denote the normal distribution draw from x
and x
and
is generated from encoding the x
together with the z
Subscene is generated from a decoder with a deep convolutional GAN architecture a by sampling on a prior z
from the latent space drawing from the previous subscenes combined with the last subscene x
. After decoding, the predict subscene is passed back to the input of the prediction model and the prior. The latent space z
is draw from
The loss of our model contains three parts, KL divergence of the prior loss , a
penalty between
and x
and an additional
penalty of the area centred around the peak of every subscene. The prediction results vary with different weight of
penalty on the peak, when the weight is too small, the model may ignore the low frequency of the subscene and
will predict the noisy small textures(crevasses) of the ice flow corresponding to
. When increasing the weight, the model predicts the peak regions but fails to generate the small textures of the ice flow.
We train our model with z and 2 LSTM layers, each layer has 128 units. By conditioning on the past eight subscenes, the results of our model on different types of subscenes are shown in Figure 3 and 4. For ice flow pattern with proper slopes(not too steep), e.g. line 2 and 6 in Figure 4, the machine learning can reproduce the slopes shapes and positions, resulting in successfully correlating two subscenes. In the experiment, the capability of reproducing small textures grows as enlarging the hidden space and batch size. However, the high pass filter’s performance differs in this two examples: in line 2, the high pass model draws the textures from
and
, since the high pass fil-ter’s results are close to binary, as long as the textures are extracted, two subscenes correlate. However, for line 6, the filter on
generates noisy signals, resulting in the failure of correlating. Another example the high pass filter fails is line 3, when the previous
does not collect the texture information(the satellite signal is affected by the cloud), in this case, the filtered subscene lacks the key information to be correlated with the filter
The machine learning model avoids the poisonous correlation subscenes using the high pass filter while over- all worse performance due to the noisy binary pixels. The machine learning model enlarges the medium correlation regions by generating continuous pixels, the peak area and learning from a range of past frames instead of just
Figure 4. The correlation map. a) persistence model(correlation between ); b) high frequency model (correlation between filter
); c) machine learning model(correlation between ml and
Table 1. Results of three models
We present IceNet dataset and encourage machine learning community to pay more attention to socially and scientifi-cally relevant datasets in the cryosphere and develop new models to help combat climate change. We also use an unsupervised learning model to predict future ice flow. Comparing to the high pass filter or persistence model, our model correlates the past and present ice flow better. Our model can also be improved if more physical and environmental parameters are introduced into the model, for example, the wind speed and the aerosol optical depth components in the atmosphere. The first parameter provides a trend for the ice flow movement and the second parameter gives us a confi-dence factor about the satellite images’ quality, dropout to particular frames can be applied if the aerosol optical depth rises over a threshold. Furthermore, black carbon aerosols were found to accelerate ice loss and glacier retreat in the Himalayas and Arctic from both wildfire soot deposition and fossil fuel emissions. Detailed analysis of the feedback effects in ’black ice’ would be a future avenue of research
The images of IceNet dataset is very different from traditional video datasets, such as in the moving-mnist, some areas are dominated by ’small textures’ while some can be smooth areas with peaks. This suggests that the transfer capability of existing models need to be investigated further or new models need to be developed for predicting the ice flow on different types of terrains around the planet.
Denton, Emily and Fergus, Rob. Stochastic video generation with a learned prior. arXiv preprint arXiv:1802.07687, 2018.
Gentine, P, Pritchard, M, Rasp, S, Reinaudi, G, and Yacalis, G. Could Machine Learning Break the Convection Parameterization Deadlock ? Geophysical Research Letters, 45:5742–5751, 2018. doi: 10.1029/2018GL078202.
IPCC. Global warming of 1.5 C. An IPCC special report on the impacts of global warming of 1.5
C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty [V. Masson-Delmotte, P. Zhai, H. O. P¨ortner, D. Roberts, J. Skea, P.R. Shukla, A. Pirani, Y. Chen, S. Connors, M. Gomis, E. Lonnoy, J. B. R. Matthews, W. Moufouma-Okia, C. P´ean, R. Pidcock, N. Reay, M. Tignor, T. Waterfield, X. Zhou (eds.)]. 2018.
Kay, J, Deser, C, Phillips, A, Mai, A, Hannay, C, Strand, G, Arblaster, J M, Bates, S C, Danabasoglu, G, Edwards, J, Holland, M, Kushner, P, Lamarque, J-F, Lawrence, D, Lindsay, K, Middleton, A, Munoz, E, Neale, R, Oleson, K, Polvani, L, and Vertenstein, M. The Community Earth System Model (CESM) Large Ensemble project. Bulletin of the American Meteorological Society, 96(8): 1333–1349, 2015. doi: 10.1175/BAMS-D-13-00255.1.
Mathieu, Michael, Couprie, Camille, Lecun, Yann, and Artificial, Facebook. Deep multi-scale video prediction beyond mean square error. arXiv preprint, 2016.
O’Gorman, Paul A and Dwyer, John G. Using machine learning to parameterize moist convection: Potential for modeling of climate, climate change, and extreme events. Journal of Advances in Modeling Earth Systems, 10(10): 2548–2563, 2018.
Rasp, Stephan, Pritchard, Michael S, and Gentine, Pierre. Deep learning to represent subgrid processes in climate models. Proceedings of the National Academy of Sciences, 115(39):1–6, 2018. doi: 10.1073/pnas. 1810286115.
Reichstein, Markus, Camps-valls, Gustau, Stevens, Bjorn, Jung, Martin, Denzler, Joachim, and Carvalhais, Nuno. Deep learning and process understanding for data-driven Earth system science. Nature, 566: 195–204, 2019. ISSN 1476-4687. doi: 10.1038/ s41586-019-0912-1. URL http://dx.doi.org/ 10.1038/s41586-019-0912-1.
Scher, Sebastian. Toward data-driven weather and climate forecasting: Approximating a simple general circulation
model with deep learning. Geophysical Research Letters, 45(22):12–616, 2018.
Schneider, Tapio, Lan, Shiwei, Stuart, Andrew, and Teixeira, Jo˜ao. Earth system modeling 2.0 : a blueprint for models that learn from observations and targeted high-resolution simulations. Geophysical Research Letters, 44:12396– 12417, 2017. doi: 10.1002/2017GL076101.
Sherwood, Steven C, Bony, Sandrine, and Dufresne, Jean- louis. Spread in model climate sensitivity traced to atmospheric convective mixing. Nature, 505:37–42, 2014. doi: 10.1038/nature12829.