IEEE-Student Branch- IIT Indore, Tech-Talks 2.0. Remote Presentation (Dec 2022) (Invited).
The popularity of machine learning methods, such as neural networks, is rapidly expanding in nearly all areas of science. The interest in these tools also coincides with a growing influx of big data and the need for high efficiency in solving predication problems. However, there is also some hesitancy in adopting the use of neural networks due to concerns about their reliability, reproducibility, and interpretability.
In climate science, we often consider signal-to-noise problems to help disentangle human-caused climate change from natural variability. These applications typically involve complicated relationships between different feedbacks at play in the ocean, cryosphere, land, and atmosphere. Recent work has shown that neural networks can be a promising tool for solving these types of statistical problems when combined with explainability techniques developed by the fields of computer science and image processing. Interestingly, these methods have revealed that neural networks often leverage regional patterns of climate change in order to make their predictions. In this webinar, I will share examples from climate science that use a few of these visualization methods to peer into the “black box” of neural networks, which help us to better understand their decision-making process while also learning new science. The same machine learning visualization methods can be easily adapted for a wide variety of applications and other scientific fields of study.
Recombination DNA Technology (Nucleic Acid Hybridization )
Machine learning for evaluating climate model projections
1. Machine learning for evaluating
climate model projections
@ZLabe
Zachary M. Labe
Postdoc at Princeton University and NOAA GFDL
7 December 2022 – Tech Talks 2.0
IEEE Student Branch ITTI & IEEE GRSS IITI
https://zacklabe.com/
5. Computer Science
Artificial Intelligence
Machine Learning
Deep Learning
Supervised
Learning
Unsupervised
Learning
Labeled data
Classification
Regression
Unlabeled data
Clustering
Dimension reduction
6. • Do it better
• e.g., parameterizations in climate models are not
perfect, use ML to make them more accurate
• Do it faster
• e.g., code in climate models is very slow (but we
know the right answer) - use ML methods to speed
things up
• Do something new
• e.g., go looking for non-linear relationships you
didn’t know were there
Very relevant for
research: may be
slower and worse,
but can still learn
something
WHY SHOULD WE CONSIDER
MACHINE LEARNING?
9. Adapted from: Kotamarthi, R., Hayhoe, K., Mearns, L., Wuebbles, D., Jacobs, J., & Jurado, J.
(2021). Global Climate Models. In Downscaling Techniques for High-Resolution Climate
Projections: From Global Change to Local Impacts (pp. 19-39). Cambridge: Cambridge University
Press. doi:10.1017/9781108601269.003
CLIMATE MODELS
Horizontal Grid
Vertical Levels
Past/Present/Future
Fully-Coupled System
20-40 Petabytes of data
15. Today’s weather or climate
scientist is far more likely to be
debugging code written in
Python… than to be poring over
satellite images or releasing
radiosondes.
“
D. Irving| Bulletin of the American Meteorological Society| 2016
16. Machine learning for weather
IDENTIFYING SEVERE THUNDERSTORMS
Molina et al. 2021
Toms et al. 2021
CLASSIFYING PHASE OF MADDEN-JULLIAN OSCILLATION
SATELLITE DETECTION
Lee et al. 2021
DETECTING TORNADOES
McGovern et al. 2019
17. Machine learning for climate
FINDING FORECASTS OF OPPORTUNITY
Mayer and Barnes, 2021
PREDICTING CLIMATE MODES OF VARIABILITY
Gordon et al. 2021
TIMING OF CLIMATE CHANGE
Barnes et al. 2019
23. E.g.,
Establish robust,
responsible AI for
severe weather
detection
Tornado Warning
Special Marine Warning
Severe Thunderstorm Warning
Flash Flood Warning
27. Artificial Neural Networks [ANN]
X1
X2
W1
W2
∑
INPUTS
NODE
Linear regression with non-linear
mapping by an “activation function”
Training of the network is merely
determining the weights “w” and
bias/offset “b"
= factivation(X1W1+ X2W2 + b)
28. Artificial Neural Networks [ANN]
X1
X2
W1
W2
∑
INPUTS
NODE
= factivation(X1W1+ X2W2 + b)
ReLU Sigmoid Linear
30. Complexity and nonlinearities of the ANN allow it to learn many
different pathways of predictable behavior
Once trained, you have an array of weights and biases which can be
used for prediction on new data
INPUT
[DATA]
PREDICTION
Artificial Neural Networks [ANN]
33. We know some metadata…
+ What year is it?
+ Where did it come from?
TEMPERATURE
34. We know some metadata…
+ What year is it?
+ Where did it come from?
TEMPERATURE
Neural network learns nonlinear
combinations of forced climate
patterns to identify the year
35. ----ANN----
2 Hidden Layers
10 Nodes each
Ridge Regularization
Early Stopping
We know some metadata…
+ What year is it?
+ Where did it come from?
[e.g., Barnes et al. 2019, 2020]
[e.g., Labe and Barnes, 2021]
TIMING OF EMERGENCE
(COMBINED VARIABLES)
RESPONSES TO
EXTERNAL CLIMATE
FORCINGS
PATTERNS OF
CLIMATE INDICATORS
[e.g., Rader et al. 2022]
Surface Temperature Map Precipitation Map
+
TEMPERATURE
36. ----ANN----
2 Hidden Layers
10 Nodes each
Ridge Regularization
Early Stopping
We know some metadata…
+ What year is it?
+ Where did it come from?
[e.g., Barnes et al. 2019, 2020]
[e.g., Labe and Barnes, 2021]
TIMING OF EMERGENCE
(COMBINED VARIABLES)
RESPONSES TO
EXTERNAL CLIMATE
FORCINGS
PATTERNS OF
CLIMATE INDICATORS
Surface Temperature Map Precipitation Map
+
TEMPERATURE
[e.g., Rader et al. 2022]
38. What is the annual mean temperature of Earth?
THE REAL WORLD
(Observations)
Anomaly is relative to 1951-1980
39. What is the annual mean temperature of Earth?
THE REAL WORLD
(Observations)
Let’s run a
climate model
40. What is the annual mean temperature of Earth?
THE REAL WORLD
(Observations)
Let’s run a
climate model
again
41. What is the annual mean temperature of Earth?
THE REAL WORLD
(Observations)
Let’s run a
climate model
again & again
42. What is the annual mean temperature of Earth?
THE REAL WORLD
(Observations)
CLIMATE MODEL
ENSEMBLES
43. What is the annual mean temperature of Earth?
THE REAL WORLD
(Observations)
CLIMATE MODEL
ENSEMBLES
Range of ensembles
= internal variability (noise)
Mean of ensembles
= forced response (climate change)
44. What is the annual mean temperature of Earth?
Range of ensembles
= internal variability (noise)
Mean of ensembles
= forced response (climate change)
But let’s remove
climate change…
45. What is the annual mean temperature of Earth?
Range of ensembles
= internal variability (noise)
Mean of ensembles
= forced response (climate change)
After removing the
forced response…
anomalies/noise!
46. What is the annual mean temperature of Earth?
• Increasing greenhouse gases (CO2, CH4, N2O)
• Changes in industrial aerosols (SO4, BC, OC)
• Changes in biomass burning (aerosols)
• Changes in land-use & land-cover (albedo)
47. What is the annual mean temperature of Earth?
• Increasing greenhouse gases (CO2, CH4, N2O)
• Changes in industrial aerosols (SO4, BC, OC)
• Changes in biomass burning (aerosols)
• Changes in land-use & land-cover (albedo)
Plus everything else…
(Natural/internal variability)
49. Greenhouse gases fixed to 1920 levels
All forcings (CESM-LE)
Industrial aerosols fixed to 1920 levels
[Deser et al. 2020, JCLI]
Fully-coupled CESM1.1
20 Ensemble Members
Run from 1920-2080
Observations
50. So what?
Greenhouse gases = warming
Aerosols = ?? (though mostly cooling)
What are the relative responses
between greenhouse gas
and aerosol forcing?
53. INPUT LAYER
HIDDEN LAYERS
OUTPUT LAYER
Surface Temperature Map
“2000-2009”
DECADE CLASS
“2070-2079”
“1920-1929”
ARTIFICIAL NEURAL NETWORK (ANN)
54. INPUT LAYER
HIDDEN LAYERS
OUTPUT LAYER
Surface Temperature Map
“2000-2009”
DECADE CLASS
“2070-2079”
“1920-1929”
BACK-PROPAGATE THROUGH NETWORK = EXPLAINABLE AI
ARTIFICIAL NEURAL NETWORK (ANN)
55. INPUT LAYER
HIDDEN LAYERS
OUTPUT LAYER
Layer-wise Relevance Propagation
Surface Temperature Map
“2000-2009”
DECADE CLASS
“2070-2079”
“1920-1929”
BACK-PROPAGATE THROUGH NETWORK = EXPLAINABLE AI
ARTIFICIAL NEURAL NETWORK (ANN)
[Barnes et al. 2020, JAMES]
[Labe and Barnes 2021, JAMES]
56. LAYER-WISE RELEVANCE PROPAGATION (LRP)
Volcano
Great White
Shark
Timber
Wolf
Image Classification LRP
https://heatmapping.org/
LRP heatmaps show regions
of “relevance” that
contribute to the neural
network’s decision-making
process for a sample
belonging to a particular
output category
Neural Network
WHY
WHY
WHY
Backpropagation – LRP
57. LAYER-WISE RELEVANCE PROPAGATION (LRP)
Volcano
Great White
Shark
Timber
Wolf
Image Classification LRP
https://heatmapping.org/
LRP heatmaps show regions
of “relevance” that
contribute to the neural
network’s decision-making
process for a sample
belonging to a particular
output category
Neural Network
WHY
WHY
WHY
Backpropagation – LRP
58. LAYER-WISE RELEVANCE PROPAGATION (LRP)
Volcano
Great White
Shark
Timber
Wolf
Image Classification LRP
https://heatmapping.org/
LRP heatmaps show regions
of “relevance” that
contribute to the neural
network’s decision-making
process for a sample
belonging to a particular
output category
Neural Network
Backpropagation – LRP
WHY
WHY
WHY
59. LAYER-WISE RELEVANCE PROPAGATION (LRP)
Image Classification LRP
https://heatmapping.org/
NOT PERFECT
Crock
Pot
Neural Network
Backpropagation – LRP
WHY
60. [Adapted from Adebayo et al., 2020]
EXPLAINABLE AI IS
NOT PERFECT
THERE ARE MANY
METHODS
61. [Adapted from Adebayo et al., 2020]
THERE ARE MANY
METHODS
EXPLAINABLE AI IS
NOT PERFECT
62. OUTPUT LAYER
Layer-wise Relevance Propagation
“2000-2009”
DECADE CLASS
“2070-2079”
“1920-1929”
BACK-PROPAGATE THROUGH NETWORK = EXPLAINABLE AI
WHY?
= LRP HEAT MAPS
[Labe and Barnes 2021, JAMES]
64. Layer-wise Relevance Propagation
BACK-PROPAGATE THROUGH NETWORK = EXPLAINABLE AI
WHY?
= LRP HEAT MAPS
Find regions of “relevance”
that contribute to the
neural network’s
decision-making process
[Labe and Barnes 2021, JAMES]
65. 1960-1999: ANNUAL MEAN TEMPERATURE TRENDS
Greenhouse gases fixed
to 1920 levels
[AEROSOLS PREVAIL]
Industrial aerosols fixed
to 1920 levels
[GREENHOUSE GASES PREVAIL]
All forcings
[STANDARD CESM-LE]
DATA
66. 1960-1999: ANNUAL MEAN TEMPERATURE TRENDS
Greenhouse gases fixed
to 1920 levels
[AEROSOLS PREVAIL]
Industrial aerosols fixed
to 1920 levels
[GREENHOUSE GASES PREVAIL]
All forcings
[STANDARD CESM-LE]
DATA
67. 1960-1999: ANNUAL MEAN TEMPERATURE TRENDS
Greenhouse gases fixed
to 1920 levels
[AEROSOLS PREVAIL]
Industrial aerosols fixed
to 1920 levels
[GREENHOUSE GASES PREVAIL]
All forcings
[STANDARD CESM-LE]
DATA
68. 1960-1999: ANNUAL MEAN TEMPERATURE TRENDS
Greenhouse gases fixed
to 1920 levels
[AEROSOLS PREVAIL]
Industrial aerosols fixed
to 1920 levels
[GREENHOUSE GASES PREVAIL]
All forcings
[STANDARD CESM-LE]
DATA
69. CLIMATE MODEL DATA PREDICT THE YEAR FROM MAPS OF TEMPERATURE
AEROSOLS
PREVAIL
GREENHOUSE GASES
PREVAIL
STANDARD
CLIMATE MODEL
[Labe and Barnes 2021, JAMES]
70. OBSERVATIONS PREDICT THE YEAR FROM MAPS OF TEMPERATURE
AEROSOLS
PREVAIL
GREENHOUSE GASES
PREVAIL
STANDARD
CLIMATE MODEL
[Labe and Barnes 2021, JAMES]
71. OBSERVATIONS
SLOPES
PREDICT THE YEAR FROM MAPS OF TEMPERATURE
AEROSOLS
PREVAIL
GREENHOUSE GASES
PREVAIL
STANDARD
CLIMATE MODEL
[Labe and Barnes 2021, JAMES]
96. N
Y
HIDDEN LAYERS
INPUT LAYER
INPUT LAYER
SAI WORLD?
or
or
map of near-surface temperature
map of near-surface temperature
map of total precipitation
map of total precipitation
Years Since
SAI Injection
OUTPUT
LOGISTIC
REGRESSION
ARTIFICAL
NEURAL
NETWORK
softmax
97. N
Y
HIDDEN LAYERS
INPUT LAYER
INPUT LAYER
SAI WORLD?
or
or
map of near-surface temperature
map of near-surface temperature
map of total precipitation
map of total precipitation
Years Since
SAI Injection
OUTPUT
LOGISTIC
REGRESSION
ARTIFICAL
NEURAL
NETWORK
softmax
110. N
Y
HIDDEN LAYERS
INPUT LAYER
INPUT LAYER
SAI WORLD?
or
or
map of near-surface temperature
map of near-surface temperature
map of total precipitation
map of total precipitation
Years Since
SAI Injection
OUTPUT
LOGISTIC
REGRESSION
ARTIFICAL
NEURAL
NETWORK
softmax
116. WE CAN LEARN NEW SCIENCE
FROM EXPLAINABLE AI.
3)
117. KEY POINTS
1. Machine learning is just another tool to add to our scientific workflow
2. We can use explainable AI (XAI) methods to peer into the black box of machine learning
3. We can learn new science by using XAI methods in conjunction with existing statistical tools
Zachary Labe
zachary.labe@noaa.gov
@ZLabe
Labe, Z.M. and E.A. Barnes (2021), Detecting climate signals using explainable AI with single-forcing
large ensembles. Journal of Advances in Modeling Earth Systems, DOI:10.1029/2021MS002464
Labe, Z.M., E.A. Barnes, and J.W. Hurrell (2023), Identifying the regional emergence of climate patterns
in a simulation of stratospheric aerosol injection, to be submitted in December 2022