Systems and methods of data preprocessing and augmentation for neural network climate forecasting models
Methods and systems for training a neural network (NN)-based climate forecasting model on a pre-processed multi-model ensemble of global climate simulation data from a plurality of global climate simulation models (GCMs), are disclosed. The methods and systems perform steps of determining a common spatial scale and a common temporal scale for the multi-model ensemble of global climate simulation data; spatially re-gridding the multi-model ensemble to the common spatial scale; temporally homogenizing the multi-model ensemble to the common temporal scale; augmenting the spatially re-gridded, temporally homogenized multi-model ensemble with synthetic simulation data generated from the spatially re-gridded, temporally homogenized multi-model ensemble; and training the NN-based climate forecasting model using the spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data. Embodiments of the present invention enable accurate climate forecasting without the need to run new dynamical global climate simulations on supercomputers.
1. A method for training a neural network (NN)-based climate forecasting model on a pre-processed multi-model ensemble of global climate simulation data from a plurality of global climate simulation models (GCMs), comprising:
determining a common spatial scale and a common temporal scale for the multi-model ensemble of global climate simulation data, wherein the multi-model ensemble of global climate simulation data comprises simulation data generated from at least two GCMs;
spatially re-gridding the multi-model ensemble of global climate simulation data to the common spatial scale;
temporally homogenizing the multi-model ensemble of global climate simulation data to the common temporal scale;
separating the spatially re-gridded, temporally homogenized multi-model ensemble of global climate simulation data into long-term signals or trends, global warming signals, seasonal signals, and other signal components caused by similar contributing climate factors;
generating synthetic simulation data through a climatology augmentation process by altering the long-term signal or trends within a feasible range;
augmenting the spatially re-gridded, temporally homogenized multi-model ensemble of global climate simulation data with the synthetic simulation data, to generate a spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data;
training, during a first training phase, the NN-based climate forecasting model using a first member time series of the spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data, wherein each input and corresponding desired output at a target lead time used in the first training phase are selected from the first member time series of the spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data; and
training, during a second training phase, the NN-based climate forecasting model using a second member time series of the spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data, wherein each input and corresponding desired output at the target lead time used in the second training phase are selected from the second member time series of the spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data;
wherein at least one training phase uses a member time series containing the synthetic simulation data.
2. The method of claim 1 , wherein the spatially re-gridding the multi-model ensemble of global climate simulation data comprises at least one of spatial interpolation and spatial extrapolation to achieve a common spatial resolution.
3. The method of claim 1 , further comprising:
computing a forecast skill score for the multi-model ensemble of global climate simulation data, based on a data predictor function, the NN-based climate forecasting model, or a model analog of the NN-based climate forecasting model; and
updating the common spatial scale based on the forecast skill score.
4. The method of claim 1 , wherein the common temporal scale comprises a common date system, and wherein the temporally homogenizing the multi-model ensemble of global climate simulation data comprises converting the multi-model ensemble of global climate simulation data to the common date system.
5. The method of claim 1 , wherein the temporally homogenizing the multi-model ensemble of global climate simulation data comprises at least one of temporal interpolation and temporal extrapolation to achieve a common temporal resolution.
6. The method of claim 1 , wherein augmenting the spatially re-gridded, temporally homogenized multi-model ensemble of global climate simulation data comprises land-sea masking.
7. The method of claim 1 , wherein the NN-based climate forecasting model is selected from the group consisting of a Convolutional Recurrent Neural Network (CRNN) having at least one Long Short-Term Memory (LSTM) layer, a Spherical Convolutional Neural Network (S 2 -CNN), and a transformer neural network.
8. The method of claim 1 , further comprising:
forecasting at least one target climate variable at the target lead time using the NN-based climate forecasting model.
9. A system for training a neural network (NN)-based climate forecasting model on a pre-processed multi-model ensemble of global climate simulation data from a plurality of global climate simulation models (GCMs), comprising:
at least one processor; and
a non-transitory physical storage medium for storing program code and accessible by the processor, the program code when executed by the processor causes the processor to:
determine a common spatial scale and a common temporal scale for the multi-model ensemble of global climate simulation data, wherein the multi-model ensemble of global climate simulation data comprises simulation data generated from at least two GCMs;
spatially re-grid the multi-model ensemble of global climate simulation data to the common spatial scale;
temporally homogenize the multi-model ensemble of global climate simulation data to the common temporal scale;
separate the spatially re-gridded, temporally homogenized multi-model ensemble of global climate simulation data into long-term signals or trends, global warming signals, seasonal signals, and other signal components caused by similar contributing climate factors;
generate synthetic simulation data through a climatology augmentation process by altering the long-term signal or trends within a feasible range;
augment the spatially re-gridded, temporally homogenized multi-model ensemble of global climate simulation data with the synthetic simulation data, to generate a spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data;
train, during a first training phase, the NN-based climate forecasting model using a first member time series of the spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data, wherein each input and corresponding desired output at a target lead time used in the first training phase are selected from the first member time series of the spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data; and
train, during a second training phase, the NN-based climate forecasting model using a second member time series of the spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data, wherein each input and corresponding desired output at the target lead time used in the second training phase are selected from the second member time series of the spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data;
wherein at least one training phase uses a member time series containing the synthetic simulation data.
10. The system of claim 9 , wherein the spatially re-grid the multi-model ensemble of global climate simulation data comprises at least one of spatial interpolation and spatial extrapolation to achieve a common spatial resolution.
11. The system of claim 9 , wherein the program code when executed by the processor further causes the processor to:
compute a forecast skill score for the multi-model ensemble of global climate simulation data, based on a data predictor function, the NN-based climate forecasting model, or a model analog of the NN-based climate forecasting model; and
update the common spatial scale based on the forecast skill score.
12. The system of claim 9 , wherein the common temporal scale comprises a common date system, and wherein the temporally homogenize the multi-model ensemble of global climate simulation data comprises converting the multi-model ensemble of global climate simulation data to the common date system.
13. The system of claim 9 , wherein the temporally homogenize the multi-model ensemble of global climate simulation data comprises at least one of temporal interpolation and temporal extrapolation to achieve a common temporal resolution.
14. The system of claim 9 , wherein the program code to augment the spatially re-gridded, temporally homogenized multi-model ensemble of global climate simulation data comprises program code for land-sea masking.
15. The system of claim 9 , wherein the NN-based climate forecasting model is selected from the group consisting of a Convolutional Recurrent Neural Network (CRNN) having at least one Long Short-Term Memory (LSTM) layer, a Spherical Convolutional Neural Network (S 2 -CNN), and a transformer neural network.
16. The system of claim 9 , wherein the program code when executed by the processor further causes the processor to:
forecast at least one target climate variable at the target lead time using the NN-based climate forecasting model.
17. A non-transitory physical storage medium for training a neural network (NN)-based climate forecasting model on a pre-processed multi-model ensemble of global climate simulation data from a plurality of global climate simulation models (GCMs), the storage medium comprising program code stored thereon, that when executed by a processor causes the processor to:
determine a common spatial scale and a common temporal scale for the multi-model ensemble of global climate simulation data, wherein the multi-model ensemble of global climate simulation data comprises simulation data generated from at least two GCMs;
spatially re-grid the multi-model ensemble of global climate simulation data to the common spatial scale;
temporally homogenize the multi-model ensemble of global climate simulation data to the common temporal scale;
separate the spatially re-gridded, temporally homogenized multi-model ensemble of global climate simulation data into long-term signals or trends, global warming signals, seasonal signals, and other signal components caused by similar contributing climate factors;
generate synthetic simulation data through a climatology augmentation process by altering the long-term signal or trends within a feasible range;
augment the spatially re-gridded, temporally homogenized multi-model ensemble of global climate simulation data with the synthetic simulation data, to generate a spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data;
train, during a first training phase, the NN-based climate forecasting model using a first member time series of the spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data, wherein each input and corresponding desired output at a target lead time used in the first training phase are selected from the first member time series of the spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data; and
train, during a second training phase, the NN-based climate forecasting model using a second member time series of the spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data, wherein each input and corresponding desired output at the target lead time used in the second training phase are selected from the second member time series of the spatially re-gridded, temporally homogenized, and augmented multi-model ensemble of global climate simulation data;
wherein at least one training phase uses a member time series containing the synthetic simulation data.
18. The non-transitory physical storage medium of claim 17 , wherein the program code when executed by the processor further causes the processor to:
forecast at least one target climate variable at the target lead time using the NN-based climate forecasting model.