Abstract The Delhi National Capital Region (NCR) receives approximately 26 days of rainfall annually, predominantly characterized by short-duration, high-intensity precipitation events. Despite the limited number of rainy days, the region frequently experiences severe waterlogging during the southwest monsoon and suffers from water scarcity during the remaining months. These issues are exacerbated by rapid urbanization and evolving hydrological dynamics. Furthermore, the spatiotemporal variability of rainfall at the local scale remains insufficiently understood, contributing to significant uncertainty in water resource management. To address this challenge, a comprehensive analysis of historical rainfall data spanning the period from 1901 to 2021 was conducted for NCR Delhi. A stochastic modelling approach, specifically the Markov chain technique, was employed to assess the probability of dry and wet weeks. Analysis of Standard Meteorological Weeks (SMWs) 27 to 39, a 12-week period revealed a high likelihood of two or three consecutive dry weeks, with probabilities ranging from 30<InlineEquation ID="IEq1"> <EquationSource Format="TEX">$\%$</EquationSource> <EquationSource Format="MATHML"><math> <mo>%</mo> </math></EquationSource> </InlineEquation> to 93<InlineEquation ID="IEq2"> <EquationSource Format="TEX">$\%$</EquationSource> <EquationSource Format="MATHML"><math> <mo>%</mo> </math></EquationSource> </InlineEquation> and 13<InlineEquation ID="IEq3"> <EquationSource Format="TEX">$\%$</EquationSource> <EquationSource Format="MATHML"><math> <mo>%</mo> </math></EquationSource> </InlineEquation> to 66<InlineEquation ID="IEq4"> <EquationSource Format="TEX">$\%$</EquationSource> <EquationSource Format="MATHML"><math> <mo>%</mo> </math></EquationSource> </InlineEquation>, respectively. Conversely, the probability of two or three consecutive wet weeks ranged from 47<InlineEquation ID="IEq5"> <EquationSource Format="TEX">$\%$</EquationSource> <EquationSource Format="MATHML"><math> <mo>%</mo> </math></EquationSource> </InlineEquation> to 98<InlineEquation ID="IEq6"> <EquationSource Format="TEX">$\%$</EquationSource> <EquationSource Format="MATHML"><math> <mo>%</mo> </math></EquationSource> </InlineEquation> and 30<InlineEquation ID="IEq7"> <EquationSource Format="TEX">$\%$</EquationSource> <EquationSource Format="MATHML"><math> <mo>%</mo> </math></EquationSource> </InlineEquation> to 91<InlineEquation ID="IEq8"> <EquationSource Format="TEX">$\%$</EquationSource> <EquationSource Format="MATHML"><math> <mo>%</mo> </math></EquationSource> </InlineEquation>, respectively. The monsoon season extends over a period of 92 days, typically from early July to the end of September, during which the total effective rainfall (ER) is estimated to be approximately 538 mm. This study underscores the importance of understanding the temporal patterns of dry and wet weeks, particularly in the context of extreme rainfall events, to support evidence-based decision-making in this critically water-stressed region. Research highlights <UnorderedList Mark="Bullet"> <ItemContent> Weekly rainfall data (1901–2021) for Delhi’s SMW 27–39 reveal significant variability, with key statistics offering insights into monsoon behavior and its impact on livelihoods. </ItemContent> <ItemContent> Markov chain modeling quantifies wet and dry week probabilities, highlighting rainfall persistence and variability during the July–September monsoon period. </ItemContent> <ItemContent> The study links rainfall trends with urbanization, groundwater depletion, and climate impacts, emphasizing sustainable water management strategies for Delhi’s long-term resilience. </ItemContent> </UnorderedList>

Rainfall analysis in Delhi NCR using stochastic modelling technique: Markov chains

期刊论文
发表日期：

2026年5月18日

查看全文

Dinkar Jha,
Anupam Gautam,
Shashi Kant

摘要

Abstract

The Delhi National Capital Region (NCR) receives approximately 26 days of rainfall annually, predominantly characterized by short-duration, high-intensity precipitation events. Despite the limited number of rainy days, the region frequently experiences severe waterlogging during the southwest monsoon and suffers from water scarcity during the remaining months. These issues are exacerbated by rapid urbanization and evolving hydrological dynamics. Furthermore, the spatiotemporal variability of rainfall at the local scale remains insufficiently understood, contributing to significant uncertainty in water resource management. To address this challenge, a comprehensive analysis of historical rainfall data spanning the period from 1901 to 2021 was conducted for NCR Delhi. A stochastic modelling approach, specifically the Markov chain technique, was employed to assess the probability of dry and wet weeks. Analysis of Standard Meteorological Weeks (SMWs) 27 to 39, a 12-week period revealed a high likelihood of two or three consecutive dry weeks, with probabilities ranging from 30 $\%$ $%$ to 93 $\%$ $%$ and 13 $\%$ $%$ to 66 $\%$ $%$ , respectively. Conversely, the probability of two or three consecutive wet weeks ranged from 47 $\%$ $%$ to 98 $\%$ $%$ and 30 $\%$ $%$ to 91 $\%$ $%$ , respectively. The monsoon season extends over a period of 92 days, typically from early July to the end of September, during which the total effective rainfall (ER) is estimated to be approximately 538 mm. This study underscores the importance of understanding the temporal patterns of dry and wet weeks, particularly in the context of extreme rainfall events, to support evidence-based decision-making in this critically water-stressed region.

Research highlights

Weekly rainfall data (1901–2021) for Delhi’s SMW 27–39 reveal significant variability, with key statistics offering insights into monsoon behavior and its impact on livelihoods.

Markov chain modeling quantifies wet and dry week probabilities, highlighting rainfall persistence and variability during the July–September monsoon period.

The study links rainfall trends with urbanization, groundwater depletion, and climate impacts, emphasizing sustainable water management strategies for Delhi’s long-term resilience.