APPLICATION OF GEOSPATIAL TECHNOLOGY FOR URBAN ENVIRONMENT STUDIES: DELHI REGION

Abstract

The urban environment encompasses of various biophysical and climatic elements, and is an ensemble of natural and built environment. With the increasing pace of urbanization, the urban environment is at substantial risk. The present work highlights two major components of urban environment viz. urban heating and urban pollution. Urban heating phenomenon is explained through its parameters land surface temperature (LST), air temperature (AT), and associated phenomena, surface urban heat island (SUHI) and atmospheric urban heat island (AUHI), quantified in terms of their intensities (SUHII and AUHII). Similarly, urban (air) pollution is explained through aerosol optical depth (AOD) and particulate matter (PM) and associated intensities of urban pollution island phenomenon (atmospheric urban pollution island or AUPI and near-surface urban pollution island or NSUPI) and related intensities (AUPII and NSUPII). The emphasis of the work revolves around the potential of geospatial technology in assessing the intensity of urban heating and urban pollution parameters over a long time period (2001 to 2018) for 53 million-plus population urban agglomerations (UA’s) (thereafter referred as cities) with special emphasis on Delhi urban region (DUR). The annual, seasonal, daytime and nighttime changes are trends are analyzed over estimated boundaries of urban conglomerates called ‘Urban’ and surrounding rural counterparts called ‘Non-urban’ using City Clustering Algorithm (CCA) over MODIS land cover datasets for each year, starting from 2001 to 2018. The changes/trends in these parameters are estimated through satellite based and ground based observations (2011 and 2018), through a non-parametric test called Mann-Kendall (MK) test. The slope of the trend is estimated using Theil Sen’s slope estimator. The analysis also encompasses the role of various climatic and anthropogenic factors in determining the urban heating and urban pollution intensities through Pearson’s correlation and stepwise multiple linear regression (MLR) for DUR. An attempt has been made to model and forecast the urban heating and urban pollution parameters using machine learning algorithm, called Artificial Neural Network (ANN) utilizing time-series forecasting module in MATLAB. To assess the ecological environmental quality (EEQ), a modified principal component analysis (PCA) based index called Remote Sensing based New Ecological Index (RSNEI) is developed using the various climatic and land based parameters for 2011 and 2018. Additionally, the urban environment was significantly affected by the pandemic (COVID-19) related lockdown in 2020 and 2021. The study determines the effect of COVID-19 linked lockdown on the urban heating and urban pollution parameters from 2017 to 2021. It was observed that the urban and non-urban areas in India, cumulatively are showing a declining urban heating phenomenon or an urban cool island (UCI) in daytime (Central India and Deccan peninsula cities with arid, semi-arid and tropical wet and dry Koppen climate system) and a significant urban heating phenomenon in the nighttime (p<0.05) at the rate of -0.032℃/yr. (p<0.05) and 0.037℃/yr. (p<0.05), respectively, for urban areas in annual daily LST values. Majority cities in the Indo Gangetic Plain show increasing nighttime summer and monsoon LST trends due to increased solar radiative heating and carbonaceous absorbing aerosols leading to positive radiative forcing effect. The AOD in urban and non-urban areas of the country is rapidly increasing (p<0.001) indicating the potential impact of vehicular emissions capturing the atmosphere in the cities (2.367%/yr.). Additionally, the AUPII (difference between AOD of urban and non-urban area) is also significantly increasing (p<0.001) in Indian cities (3.599%/yr.). Delhi Urban Region (DUR) represents intense nighttime heating rate in LST for both urban and non-urban areas (p<0.05). and a non-significant SUHII heating rate from 2001 to 2018. The average annual SUHIINight>SUHIIDay by annual daily SUHII of 2.139℃ over 18 years’ time-period. Similar increasing trends (p<0.05) are observed for AOD and AUPII. The most important factors affecting the urban heating assessed using Pearson’s correlation and MLR were nighttime lights (NL), enhanced vegetation index (EVI), air temperature (AT), evapotranspiration (ET) and population density (PD), whereas, white sky albedo (WSA), ET, NL and EVI affected the SUHII intensely. AOD has a negative influence on LST, inspite of having a high positive correlation with it during daytime. AOD mostly plays a negative radiative forcing during the day and the relationship is non-linear and not well explained by stepwise MLR or Pearson’s correlation. AUPII effect is clearly understood by ΔWSA, ΔPD, urbanization rate (ΔUR), ΔNL, ΔEVI and ΔLST. The continuous increasing values in LST have shown a sudden peak in 2016 due to heat wave as an aftermath of El-Nino experienced during the same time and a considerable reduction in AOD was observed after 2016 due to the introduction of government interventions towards cleaner fuel consumption. Using satellite and ground based measurements for 2011 and 2018, it was ascertained that 2018 year has seen reduction in terms of annual and seasonal SUHII and AUHII values. On the contrary, nighttime SUHII and AUPII mean values have seen a considerable increase. An extensive reduction in AUPII and NSUPII values are also observed from 2011 to 2018. Supporting the above findings, the nighttime RSNEI maps also explain the deteriorating environmental regime especially during the nighttime in 2018 compared to 2011, where LST and normalized difference soil index (NDSI) as heating and dryness indicators, respectively, have shown a negative influence on EEQ, whereas, NDVI (vegetation) and wetness indicate a positive influence on EEQ. AOD has shown a minimal positive influence on daytime and negative influence on nighttime EEQ highlighting the role of negative radiative forcing effect of aerosols on urban micro-climate. It was also observed that impervious surface area (ascertained using normalized difference anthropogenic impervious surface area index, NDAISI) has a positive correlation with LST (R2>0.94) during nighttime in the DUR, indicating that nighttime gives a better understanding of SUHI phenomenon, aggravating due to urbanization. ANN was successful in forecasting LST, AOD, PM and AT scenarios for 2024 and 2028 with RMSE of 0.01 and R2=0.92 sing various climatic and anthropogenic land based variables. Additionally, it was deciphered that COVID-19 related lockdown improved the environmental regime in terms of LST, NO2 and AOD in 2020 compared to 2016-19. The SUHII, AUPII and NSUPII values reduced during 2020 as a result of constrained anthropogenic activities. The research work highlights the role of geospatial technology (remote sensing and GIS) in understanding scenarios of urban micro-climate indicators and paves way for policy planning interventions leading to sustainable environment in an urban purlieu.