Unsupervised Discovery of El Nino Using Causal Feature Learning on Microlevel Climate Data
نویسندگان
چکیده
We show that the climate phenomena of El Niño and La Niña arise naturally as states of macrovariables when our recent causal feature learning framework (Chalupka et al., 2015, 2016) is applied to micro-level measures of zonal wind (ZW) and sea surface temperatures (SST) taken over the equatorial band of the Pacific Ocean. The method identifies these unusual climate states on the basis of the relation between ZW and SST patterns without any input about past occurrences of El Niño or La Niña. The simpler alternatives of (i) clustering the SST fields while disregarding their relationship with ZW patterns, or (ii) clustering the joint ZW-SST patterns, do not discover El Niño. We discuss the degree to which our method supports a causal interpretation and use a low-dimensional toy example to explain its success over other clustering approaches. Finally, we propose a new robust and scalable alternative to our original algorithm (Chalupka et al., 2016), which circumvents the need for high-dimensional density learning.
منابع مشابه
El Niño Southern Oscillation and Primary Agricultural Commodity Prices: Causal Inferences from Smooth Transition Models
Global climate anomalies affect world economies and primary commodity prices. One of the more pronounced climate anomalies is El Niño Southern Oscillation (ENSO). In this study I examine the relationship between ENSO and world commodity prices using monthly time series of the sea-surface temperature anomalies in the Nino 3.4 region, and real prices of thirty primary agricultural commodities. I ...
متن کاملClimate Change Prediction Using Data Mining
Great progress has been made in the effort to understand and predict El Nino, the anomalous warming of the sea surface temperature (SST) along the equator off the coast of South America which has a strong impact on the climate change over the world. Advances in improved climate predictions will result in significantly enhanced economic opportunities, particularly for the national agriculture, f...
متن کاملBook Review: Computational Methods of Feature Selection
Feature selection selects a subset of relevant features, and also removes irrelevant and redundant features from the data to build robust learning models. Feature selection is very important, not only because of the curse of dimensionality, but also due to emerging data complexities and quantities faced by multiple disciplines, such as machine learning, data mining, pattern recognition, statist...
متن کاملانسو و رخدادهای سالانه لیشمانیوز جلدی در ایران
Background and Aims: Cutaneous leishmaniosis (CL) is a neglected tropical vector-borne disease. As other vector-transmitted diseases, its transmission is sensitive to the physical environment. The purpose of this study was to investigate the relationship between El Niño-Southern Oscillation (ENSO) and annual incidence of CL in Iran. Materials and Methods: The data of ENSO were obtained from ...
متن کاملClimate changes over the past millennium: Relationships with Mediterranean climates(∗)
— Evidence is reviewed for climate change and its causes over the interval spanning roughly the past millennium. Particular emphasis is placed on patterns of climate change influencing Mediterranean climates of the Northern Hemisphere. The evidence is taken from studies using high-resolution climate “proxy” data sources, and climate modeling simulations. The available evidence suggests that for...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- CoRR
دوره abs/1605.09370 شماره
صفحات -
تاریخ انتشار 2016