Abstract
Projection of changes in extreme indices of climate variables such as temperature and precipitation are critical to assess the potential impacts of climate change on human-made and natural systems, including critical infrastructures and ecosystems. While impact assessment and adaptation planning rely on high-resolution projections (typically in the order of a few kilometers), state-of-the-art Earth System Models (ESMs) are available at spatial resolutions of few hundreds of kilometers. Current solutions to obtain high-resolution projections of ESMs include downscaling approaches that consider the information at a coarse-scale to make predictions at local scales. Complex and non-linear interdependence among local climate variables (e.g., temperature and precipitation) and large-scale predictors (e.g., pressure fields) motivate the use of neural network-based super-resolution architectures. In this work, we present auxiliary variables informed spatio-temporal neural architecture for statistical downscaling. The current study performs daily downscaling of precipitation variable from an ESM output at 1.15 degrees (115 km) to 1/4 degrees (25 km) over one of the most climatically diversified countries, India. We showcase significant improvement gain against two popular state-of-the-art baselines with a better ability to predict statistics of extreme events. To facilitate reproducible research, we make available all the codes, processed datasets, and trained models in the public domain.
Highlights
The last few decades have witnessed record-breaking climate and weather-related extremes across the globe [1], [2]
Schemes to account for interactions across various constituting components, Earth system models (ESMs) suffer from three broad sources of uncertainties: (i) knowledge gaps in understanding of coupled natural-human systems which could results in different Representative Concentration Pathways (RCPs), (ii) lack of understanding of physics of climate system which is encapsulated through parametric differences in Multiple Model Ensembles (MMEs), and (iii) intrinsic variability which is captured through multiple initial condition ensembles [6]
We present a Recurrent Convolutional Long Short Term Memory (LSTM) based Super-resolution approach towards statistical downscaling of climatic data from coarse-resolution ESM to fine-resolution Observation data, which accounts for spatial and temporal dependence in space and time between target variable and auxiliary variables
Summary
The last few decades have witnessed record-breaking climate and weather-related extremes across the globe [1], [2]. Schemes to account for interactions across various constituting components, ESMs suffer from three broad sources of uncertainties: (i) knowledge gaps in understanding of coupled natural-human systems which could results in different Representative Concentration Pathways (RCPs), (ii) lack of understanding of physics of climate system which is encapsulated through parametric differences in Multiple Model Ensembles (MMEs), and (iii) intrinsic variability which is captured through multiple initial condition ensembles [6]. To characterize these uncertainties, ESMs are often executed with multiple initial conditions, multiple ensemble mode with different RCPs making impact relevant insight discovery from ESM outputs a ‘‘big data’’ challenge [7]. While there has been an emphasis on multiple model ensembles for assessing climate change impacts on
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
