Classification of summer monsoon rainfall patterns over India

The East Asian Summer Monsoon (EASM) is a crucial monsoon system that profoundly influences the summer climate in Southeast China (SEC). Classification of monsoon rainfall patterns is vital to physical diagnosis, rainfall prediction and identification of sites that are prone to rainfall-triggering floods. With the great endeavors on understanding the complexity of the EASM in the past decades, the traditionally accepted rainfall patterns in SEC and the relevant analyses appear outdated or even inadequate. Having highly-improved observations at hand helps update the monsoon rainfall patterns in SEC and the potential predictability.The present study employs a nonlinear neural network classification technique, the Self-organizing map (SOM), to identify the rainfall patterns in SEC based on gauge data. Three distinct rain belts over the Huai River basin (HRB), lower Yangtze River basin (LYRB) and South Coast region (SCR) are found. Their subseasonal variability highly agrees with the stepwise progression of the East Asian Summer Monsoon (EASM) front in space and time. Analysis reveals that precipitation in the SCR and HRB rain belts undergo a regime shift after the mid-1990s, whereas the 1990s is the most active decade for the LYRB rain belt. These systematic changes are in abreast with similar changes in EASM and other climate events documented in the literature.Additionally, a SOM-based algorithm is developed to further divide gauge stations into three groups featuring homogeneous rain belt patterns. Promising predictability of group-averaged daily rainfall is then achieved, with about 39% to 50% of the total variance explained by circulation-informed regression models, verified by both cross-validation and blind prediction. Through further diagnosis in the useful predictors, the western North Pacific subtropical high, blocking high anomalies over northeast China and the upper-level divergence over SEC, are found to best explain the variability of the rain belts. The proposed Russia-China wave pattern (western/central Russia &#8594; north of Tibetan Plateau &#8594; SEC) and teleconnection between the El Ni&#241;o-Southern Oscillation and the rain belts also offer additional predictability. This study aims to set an updated benchmark on the summer monsoon rainfall patterns in SEC, from which the promising daily predictability and the informative circulation patterns are obtained. Findings from this work may also advance the understanding of the EASM rain belts, and offer insights to the source of bias for numerical simulations of daily summer monsoon rainfall in the region.

Download Full-text

Periodicities and ENSO relationships of the seasonal precipitation over six major sub-divisions of India

MAUSAM ◽

10.54302/mausam.v50i1.1803 ◽

2021 ◽

Vol 50 (1) ◽

pp. 43-54

Author(s):

R. P. KANE

Keyword(s):

Time Series ◽

Spectral Analysis ◽

Summer Monsoon ◽

Monsoon Rainfall ◽

Summer Monsoon Rainfall ◽

Seasonal Precipitation ◽

North Central ◽

India Summer Monsoon ◽

India Summer Monsoon Rainfall

A spectral analysis of the 1848-1995 (148 year) time series of Sontakke and Singh (1996) representing estimates of summer monsoon (June-September) precipitation amounts over six homogeneous zones (Northwest NW, North central NC, Northeast NE. West Peninsular WP, East Peninsular EP, South Peninsular SP) and the whole of India (AI) revealed significant periodicities in the QBO and QTO regions (2-3 years and 3-4 years) as also higher periodicities, some common to all zones. To study the ENSO relationship, a finer classification of years was adopted. For the All India summer monsoon rainfall as also for all the zones except NE, Unambiguous ENSOW (where El Nino existed and SOI minima and SST maxima were in the middle of the calendar year i.e., May-August), were overwhelmingly associated with droughts and the cold (C) events were associated with floods. For other types of events, the results were uncertain and a few extreme rainfalls occurred even during some Non-events.

Download Full-text

Binary classification of rainfall time-series using machine learning algorithms

International Journal of Electrical and Computer Engineering (IJECE) ◽

10.11591/ijece.v12i2.pp1945-1954 ◽

2022 ◽

Vol 12 (2) ◽

pp. 1945

Author(s):

Shilpa Hudnurkar ◽

Neela Rayavarapu

Keyword(s):

Summer Monsoon ◽

Monsoon Rainfall ◽

Binary Classification ◽

Summer Monsoon Rainfall ◽

The State ◽

Misclassification Rate ◽

Annual Rainfall ◽

Support Vector ◽

Weather Variables

Summer monsoon rainfall contributes more than 75% of the annual rainfall in India. For the state of Maharashtra, India, this is more than 80% for almost all regions of the state. The high variability of rainfall during this period necessitates the classification of rainy and non-rainy days. While there are various approaches to rainfall classification, this paper proposes rainfall classification based on weather variables. This paper explores the use of support vector machine (SVM) and artificial neural network (ANN) algorithms for the binary classification of summer monsoon rainfall using common weather variables such as relative humidity, temperature, pressure. The daily data, for the summer monsoon months, for nineteen years, was collected for the Shivajinagar station of Pune in the state of Maharashtra, India. Classification accuracy of 82.1 and 82.8%, respectively, was achieved with SVM and ANN algorithms, for an imbalanced dataset. While performance parameters such as misclassification rate, F1 score indicate that better results were achieved with ANN, model parameter selection for SVM was less involved than ANN. Domain adaptation technique was used for rainfall classification at the other two stations of Maharashtra with the network trained for the Shivajinagar station. Satisfactory results for these two stations were obtained only after changing the training method for SVM and ANN.

Download Full-text