Publications

Geospatial Diffusion for Land Cover Imperviousness Change Forecasting

Published in Proceedings of the 33rd ACM International Conference on Advances in Geographic Information Systems , 2025

Land-use and land-cover (LULC) has a significant effect on several Earth system processes. For example, impervious surfaces reduce infiltration and speed water flow, impacting regional hydrology and flood risk. While Earth System models have improved forecasting hydrologic and atmospheric processes at higher resolutions, the ability to forecast LULC change has lagged behind. In this paper, we propose a new paradigm exploiting Generative AI (GenAI) for land cover change forecasting by framing it as a data synthesis problem conditioned on historical and auxiliary data-sources. To demonstrate the feasibility of our methodology, we perform experiments where a diffusion model is trained for decadal forecasting of imperviousness change across the entire United States. We find that our model yields MAE lower than a no-change baseline for resolutions ≥ 0.7 X 0.7km2 on average, demonstrating its ability to capture and project accurate spatiotemporal patterns. Finally, we discuss future research to incorporate Earth’s physical properties and enabling scenario simulations via driver variables.

Download here

Rising infrastructure inequalities accompany urbanization and economic development

Published in Nature Communications , 2025

Impending global urban population growth is expected to occur with considerable infrastructure expansion. However, our understanding of attendant infrastructure inequalities is limited, highlighting a critical knowledge gap in the sustainable development implications of urbanization. Using satellite data from 2000 to 2019, we examine country-level population-adjusted biases in infrastructure distribution within and between regions of varying urbanization levels and derive four key findings. First, we find long-run positive associations between infrastructure inequalities and both urbanization and economic development. Second, our estimates highlight increasing infrastructure inequalities across most of the countries examined. Third, we find greater future infrastructure inequality increases in the global south, where inequalities will rise more in countries with substantial urban primacy. Fourth, we find that infrastructure inequality may evolve differently than economic inequalities. Overall, advancing sustainable development vis-à-vis urbanization and economic development will require intentional infrastructure planning for spatial equity.

Download here

Urbanization and malaria have a contextual relationship in endemic areas: A temporal and spatial study in Ghana

Published in PLOS Global Public Health , 2024

In West Africa, malaria is one of the leading causes of disease-induced deaths. Existing studies indicate that as urbanization increases, there is corresponding decrease in malaria prevalence. However, in malaria-endemic areas, the prevalence in some rural areas is sometimes lower than in some peri-urban and urban areas. Therefore, the relationship between the degree of urbanization, the impact of living in urban areas, and the prevalence of malaria remains unclear. This study explores this association in Ghana, using epidemiological data at the district level (2015–2018) and data on health, hygiene, and education. We applied a multilevel model and time series decomposition to understand the epidemiological pattern of malaria in Ghana. Then we classified the districts of Ghana into rural, peri-urban, and urban areas using administratively defined urbanization, total built areas, and built intensity. We converted the prevalence time series into cross-sectional data for each district by extracting features from the data. To predict the determinant most impacting according to the degree of urbanization, we used a cluster-specific random forest. We find that prevalence is impacted by seasonality, but the trend of the seasonal signature is not noticeable in urban and peri-urban areas. While urban districts have a slightly lower prevalence, there are still pockets with higher rates within these regions. These areas of high prevalence are linked to proximity to water bodies and waterways, but the rise in these same variables is not associated with the increase of prevalence in peri-urban areas. The increase in nightlight reflectance in rural areas is associated with an increased prevalence. We conclude that urbanization is not the main factor driving the decline in malaria. However, the data indicate that understanding and managing malaria prevalence in urbanization will necessitate a focus on these contextual factors. Finally, we design an interactive tool, ’malDecision’ that allows data-supported decision-making.

Download here

Toward Zero-Carbon Urban Transitions with Health, Climate Resilience, and Equity Co-Benefits: Assessing Nexus Linkages

Published in Annual Review of Environment and Resources , 2023

Getting to net-zero-carbon cities while advancing well-being (W), health (H), social equity (E), and climate resilience (R) (referred to as the WHER outcomes) is critical for local and global sustainability. However, science is nascent on the linkages between zero-carbon pathways and WHER outcomes. This article presents a transboundary urban metabolism framework, rooted in seven key infrastructure and food provisioning systems, to connect urban decarbonization strategies with WHER outcomes. Applying the framework along with a literature review, we find the evidence for co-beneficial decarbonization to be strong for health; limited for well-being; uncertain for resilience; and requiring intentional design to advance equity, including distributional, procedural, and recognitional aspects. We describe the evidence base, identify key knowledge gaps, and delineate broad parameters of a new urban nexus science to enable zero-carbon urban transitions with WHER co-benefits. We highlight the need for fine-scale data encompassing all seven sectors across scales, along with multiple and multiscale climate risks, accompanied by next-generation multisector, multiscale, multioutcome nexus models.

Download here

Integrating Ecological and Social Concepts for Urban Metabolism Studies

Published in Urban Metabolism and Climate Change, Springer Nature Link , 2023

Urban sustainability efforts are increasingly related to concepts from self-sufficient natural ecosystems. Focusing on the metabolic characteristics and pathways of urban areas from an ecological lens can help advance our understanding of how they can be sustainable. The ecological lens allows the simultaneous notion of an urban area as a system and as a natural entity and offers insights into fundamental mechanisms that help identify natural solutions over purely technological ones. This chapter develops a systematic understanding of how urban metabolism aligns with ecosystem concepts and how social dimensions, and associated multi-dimensional and multi-scalar inequalities, add novel characteristics, tradeoffs, and synergies. In doing so, we develop an analytical framework that provides an outlook for sustainability goals relevant to urban areas including developing ecologically informed solutions and improving resource efficiency in urban areas.

Download here

Detecting subpixel human settlements in mountains using deep learning: A case of the Hindu Kush Himalaya 1990–2020

Published in Remote Sensing of Environment , 2023

The majority of future population growth in mountains will occur in small- and medium-sized cities and towns and affect vulnerable ecosystems. However, mountain settlements are often omitted from global land cover analyses due to the low spatial resolution of satellite images, which cannot resolve the small scale of mountains settlements. This study demonstrates, for the first time, the potential of deep learning to detect human settlements in mountains at the sub-pixel level, based on Landsat satellite imagery. We hypothesized that adding spatial and temporal features could improve the detection of mountain settlements since spectral information alone led to inaccurate results. For spatial features, we compared a U-shaped neural network (U-Net), a deep learning algorithm that automatically learns spatial features, with a simple random forest (RF) algorithm. Then, we assessed whether temporal features would increase accuracy by comparing two input datasets, multispectral imagery and temporal features from the Continuous Change Detection and Classification (CCDC) algorithm. We evaluated each method by calculating the accuracies of (1) the binary settlement footprint, (2) the subpixel estimates of impervious surfaces, and (3) urban growth. We tested the accuracies using visually interpreted datasets from time-series Google Earth images across the Hindu Kush Himalaya that were not used for training to evaluate model transferability. The U-Net successfully improved mountain settlement mapping compared to the random forest, with a substantial discrepancy in small settlements. The time-series results from the U-Net successfully captured long-term urban growth but fewer short-term changes. Contrary to expectations, the CCDC temporal features reduced the accuracy of mountain settlement mapping due to frequent cloud cover in hilly areas. Our subpixel analysis reveals that the built-up area of the Hindu Kush Himalaya has expanded at a rate of 61 km2 per year from 1990 to 2020, which is about twice the estimate of the Global Human Settlement Layer using binary urban/non-urban classifications.

Download here

Characterizing COVID-19 waves in urban and rural districts of India

Published in npj Urban Sustainability , 2022

Understanding spatial determinants, i.e., social, infrastructural, and environmental features of a place, which shape infectious disease is critically important for public health. We present an exploration of the spatial determinants of reported COVID-19 incidence across India’s 641 urban and rural districts, comparing two waves (2020–2021). Three key results emerge using three COVID-19 incidence metrics: cumulative incidence proportion (aggregate risk), cumulative temporal incidence rate, and severity ratio. First, in the same district, characteristics of COVID-19 incidences are similar across waves, with the second wave over four times more severe than the first. Second, after controlling for state-level effects, urbanization (urban population share), living standards, and population age emerge as positive determinants of both risk and rates across waves. Third, keeping all else constant, lower shares of workers working from home correlate with greater infection risk during the second wave. While much attention has focused on intra-urban disease spread, our findings suggest that understanding spatial determinants across human settlements is also important for managing current and future pandemics.

Download here

Satellite Data Applications for Sustainable Energy Transitions

Published in Frontiers in Sustainability , 2022

The transition to a sustainable energy system will be one of the greatest challenges of the coming decade and beyond. New sources of information that can support decision-making are essential for this transition. A growing portfolio of satellite data products offer insights to support energy policy and planning. Here we present a review of satellite data applications in four key areas related to energy: supply, demand, impacts, and resilience. Satellite data provide greater spatial and temporal coverage in areas where other data are scarce and can complement other information sources to provide a more complete picture of the global energy system. We find that satellite data are already being applied to a wide range of energy topic areas with varying needs, from planning and operation of renewable energy projects, to tracking changing patterns in energy access and use, to monitoring environmental impacts and verifying the effectiveness of emissions reduction efforts. While satellite data can play an increased role throughout the policy and planning lifecycle, there are technical, social, and structural barriers to increased use. We conclude with a discussion of opportunities for satellite data applications to energy and recommendations for research to maximize the value of satellite data for sustainable energy transitions.

Download here

Infrastructure Inequality is a Characteristic of Urbanization

Published in Proceedings of the National Academy of Sciences , 2022

Urbanization can challenge sustainable development if it produces unequal outcomes. Infrastructure is an important urbanization dimension, providing services to support diverse urban activities. However, it can lock in unequal outcomes due to its durable nature. This paper studies inequalities in infrastructure distributions to derive insights into the structure and characteristics of unequal outcomes associated with urbanization. We analyzed infrastructure inequalities in two emerging economies in the Global South: India and South Africa. We developed and applied an inequality measure to understand the structure of inequality in infrastructure provisioning (based on census data) and infrastructure availability (based on satellite nighttime lights [NTLs] data). Consistent with differences in economic inequality, results show greater inequalities in South Africa than in India and greater urban inequalities than rural inequalities. Nevertheless, inequalities in urban infrastructure provisioning and infrastructure availability increase from finer to coarser spatial scales. NTL-based inequality measurements additionally show that inequalities are more concentrated at coarse spatial scales in India than in South Africa. Finally, results show that urban inequalities in infrastructure provisioning covary with urbanization levels conceptualized as a multidimensional phenomenon, including demographic, economic, and infrastructural dimensions. Similarly, inequalities in urban infrastructure availability increase monotonically with infrastructure development levels and urban population size. Together, these findings underscore infrastructure inequalities as a feature of urbanization and suggest that understanding urban inequalities requires applying an inequality lens to urbanization.

Download here

Urbanization and food consumption in India

Published in Scientific Reports , 2020

The shift towards urban living is changing food demand. Past studies on India show significant urban–rural differences in food consumption. However, a scientific understanding of the underlying relationships between urbanization and food consumption is limited. This study provides the first detailed analysis of how urbanization influences both quantity and diversity of food consumption in India by harnessing the strength of multiple datasets, including consumer expenditure surveys, satellite imagery, and census data. Our statistical analysis shows three main findings. First, in contrast to existing studies, we find that much of the variation in food consumption quantity is due to income and not urbanization. After controlling for income and state-level differences, our results show that average consumption is higher in urban than rural areas for fewer than 10% of all commodities. That is, there is nearly no difference in average consumption between urban and rural residents. Second, we find the influence of urbanization as a population share on food consumption diversity to be statistically insignificant (p-value > 0.1). Instead, the results show that infrastructure, market access, percentage working women in urban areas, and norms and institutions have a statistically significant influence. Third, all covariates of food consumption diversity we tested were found to be associated with urbanization. This suggests that urbanization influences on food consumption are both indirect and multidimensional. These results show that increases in the urban population size alone do not explain changes in food consumption in India. If we are to understand how food consumption may change in the future due to urbanization, the study points to the need for a more complex and multidimensional understanding of the urbanization process that goes beyond demographic shifts.

Download here

Urbanization, processed foods, and eating out in India

Published in Global Food Security , 2020

Urban consumption of processed and fast foods is a challenge to nutrition security. Observed differences in urban versus rural consumption are commonly attributed to higher income levels in urban areas. Yet, there is still no clear understanding of why and how urban dwellers consume differently. Using India as a case study, we analyze expenditures on processed foods and consumption of food away from home (FAFH) of urban, metropolitan, and rural populations using OLS regression models. We show that urban households spend more on processed foods and consume more FAFH than rural households. Most of this difference can be attributed to differing socio-economic and demographic factors, such as higher income, or smaller urban household size. However, even after controlling for these factors, we find differences not only between rural and urban areas but also between different urban areas: households in large metropolitan areas consume more than households in smaller non-metropolitan urban areas. These inter-urban variations suggest that the dichotomy of urban versus rural consumption does not adequately capture the full spectrum of food consumption complexities. Our findings indicate that urbanization is affecting how people consume food beyond shaping their socio-economic and demographic status. We also highlight the need to account for the role of urbanization—beyond an urban-rural dichotomy—when addressing the challenges associated with changing food consumption patterns.

Download here

Urban Land Use: Central to Building a Sustainable Future

Published in One Earth , 2019

Urban land use has impacts spread across time and place. How urban areas are built and configured in the coming decades will have long-term implications on the environment and the lives of billions of people. A systems approach that emphasizes the importance of science and the relevance of policy to understanding urban land use is vital to building a sustainable future.

Download here

Modelling Spatial Patterns of Urban Growth in Pune Metropolitan Region, India

Published in Applications and Challenges of Geospatial Technology , 2018

Explaining urban growth patterns is a fundamental need to understand the recent rapid urbanization globally. This study identifies geographic features explaining the spatial patterns of urban land expansion (ULE) in the rapidly urbanizing Pune metropolitan region (India). ULE maps were derived from Landsat Thematic Mapper and Operational Land Imager images using support vector machine (SVM) classification. Relation between geographic features and spatial patterns of ULE was analyzed using statistical modelling including ordinary least squares (OLS) regression, spatial lag model (SLM), spatial error model (SEM), and geographically weighted regression (GWR). SEM specification best modeled ULE patterns. High density of existing urban areas is identified to negatively affect ULE, suggesting dominant dispersed urban growth. In addition, proximity to special economic zones and transportation infrastructure explains multicentric growth in the region. GWR model was identified inappropriate due to the presence of high local collinearity. Models accounting for spatial dependencies are recommended while studying ULE patterns.

Download here

Time series analysis of satellite data to characterize multiple land use transitions: a case study of urban growth and agricultural land loss in India

Published in Journal of Land Use Science , 2018

Urban land use change is one of the most impactful land transitions on the biosphere, resulting in land conversion, habitat loss, and changes in biogeochemical cycling, climate, and hydrology. Thus, understanding it is essential for global change research. Most land change detection algorithms assume linear changes. However, urban land-use changes are often nonlinear, i.e., follow multiple transitions over time. We propose a new methodology to identify multiple transitions due to urbanization with high frequency remote sensing time series. We design, implement, and evaluate a time series approach to detect the timing of urban conversion of agricultural land in India. Results show an overall accuracy of 82.11% in detecting change timing when the algorithm is applied to MODIS normalized difference vegetation index (NDVI) time series. The proposed algorithm yields better results with raw time series than filtered time series. We discuss the usefulness of our algorithm to understand nonlinear land transitions.

Download here

Monitoring Annual Vegetated Land Loss to Urbanization with Landsat Archive: A Case Study in Shanghai, China

Published in Remote Sensing Time Series Image Processing , 2018

Urban expansion (urbanization) often causes significant disturbance to ecosystems surrounding cities, sometimes resulting in the removal of large amounts of biomass and in turn putting the human-nature systems at risk. Landsat imagery has long been utilized to monitor urbanization and ecosystem change at regional and local scales. However, few studies use Landsat time series to monitor urbanization at higher temporal frequencies, especially for applications focusing on large geographic areas, mainly due to the lack of efficient algorithms and computation facilities to handle large data volume. Here we extract annual vegetated land loss to urbanization information with Landsat time series and implement it on the Google Earth Engine (GEE), a cloud-computing platform, for large area applications. We first generate annual Landsat cloud/shadow free NDVI mosaics and then NDVI time series spanning the period from 2000 to 2010. We then develop change and stable models to identify change time points in the time series. We evaluate the performance of the proposed method in Shanghai, China, which has experienced rapid urbanization during the past few decades. Results show annual ecosystem disturbance caused by urban expansion is well captured, with a change detection accuracy higher than 80%. Our method is fast, simple, and can be easily extended to large areas on the Google Earth Engine cloud-computing platform.

Download here

NASA’s Black Marble nighttime lights product suite

Published in Remote Sensing of Environment , 2018

NASA’s Black Marble nighttime lights product suite (VNP46) is available at 500 m resolution since January 2012 with data from the Visible Infrared Imaging Radiometer Suite (VIIRS) Day/Night Band (DNB) onboard the Suomi National Polar-orbiting Platform (SNPP). The retrieval algorithm, developed and implemented for routine global processing at NASA’s Land Science Investigator-led Processing System (SIPS), utilizes all high-quality, cloud-free, atmospheric-, terrain-, vegetation-, snow-, lunar-, and stray light-corrected radiances to estimate daily nighttime lights (NTL) and other intrinsic surface optical properties. Key algorithm enhancements include: (1) lunar irradiance modeling to resolve non-linear changes in phase and libration; (2) vector radiative transfer and lunar bidirectional surface anisotropic reflectance modeling to correct for atmospheric and BRDF effects; (3) geometric-optical and canopy radiative transfer modeling to account for seasonal variations in NTL; and (4) temporal gap-filling to reduce persistent data gaps. Extensive benchmark tests at representative spatial and temporal scales were conducted on the VNP46 time series record to characterize the uncertainties stemming from upstream data sources. Initial validation results are presented together with example case studies illustrating the scientific utility of the products. This includes an evaluation of temporal patterns of NTL dynamics associated with urbanization, socioeconomic variability, cultural characteristics, and displaced populations affected by conflict. Current and planned activities under the Group on Earth Observations (GEO) Human Planet Initiative are aimed at evaluating the products at different geographic locations and time periods representing the full range of retrieval conditions.

Download here

Mapping floods and assessing flood vulnerability for disaster decision-making: a case study remote sensing application in Senegal

Published in Earth Observation Open Science and Innovation , 2018

Download here

Comparative evaluation of relative calibration methods for DMSP/OLS nighttime lights

Published in Remote Sensing of Environment , 2017

With over two decades of historical observations, DMSP/OLS nighttime lights (NTL) are an invaluable asset for monitoring, characterizing, and understanding human activity. Due to the lack of on-board calibration, there are systematic biases in NTL data. Consequently, a key deterrent to the use of the entire NTL archive is the difficulty in generating a consistent NTL time series. Currently, there are a number of methods to calibrate NTL data in order to generate a consistent time series. However, there is no systematic evaluation of the performances of these calibration efforts. The purpose of this paper is to compare and evaluate existing calibration methods and to provide guidance to NTL end users on their relative strengths and weaknesses. We apply two widely adopted criteria to assess the performance of the methods: a) systematic bias minimization, and b) relation with ancillary socio-economic data on economic activity and electricity consumption. Our results show that global-scale calibration methods outperform regionally-based calibration methods. Furthermore, inconsistencies, in the form of differences in pixel values from two satellites of the same region and year, continue to exist in NTL data even after the calibration methods are applied, both across regions and across scales. We find that dimly lit regions are more difficult to calibrate compared to brightly lit regions and that inconsistencies persist in higher northern and lower southern latitudes. We conclude that significant improvements in calibration results will require a shift either towards absolute correction methods using in-situ data or relative calibration methods using NTL from VIIRS.

Download here

A Robust Method to Generate a Consistent Time Series From DMSP/OLS Nighttime Light Data

Published in IEEE Transactions on Geoscience and Remote Sensing , 2016

The long time series of nighttime light (NTL) data collected by DMSP/OLS sensors provides a unique and valuable resource to study changes in human activities. However, its full time series potential has not been fully explored, mainly due to inconsistencies in the temporal signal. Previous studies have tried to resolve this issue in order to generate a consistent NTL time series. However, due to geographic limitations with the algorithms, these approaches cannot generate a coherent NTL time series globally. The purpose of this study is to develop a methodology to create a consistent NTL time series that can be applied globally. Our method is based on a novel sampling strategy to identify pseudoinvariant features. We select data points along a ridgeline-the densest part of a density plot generated between the reference image and the target image-and then use those data points to derive calibration models to minimize inconsistencies in the NTL time series. Results show that the algorithm successfully calibrates DMSP/OLS annual composites and generates a consistent NTL time series. Evaluation of the results shows that the calibrated NTL time series significantly reduces the differences between two images within the same year and increases the correlations between the NTL time series and gross domestic product as well as with energy consumption, and outperforms the Elvidge et al. (2014) method. The methodology is simple, robust, and easy to implement. The quality-enhanced NTL time series can be used in a myriad of applications that require a consistent data set over time.

Download here

Contemporary urbanization in India

Published in The Routledge Handbook of Urbanization and Global Environmental Change , 2015

Download here

Numerical modelling spatial patterns of urban growth in Chandigarh and surrounding region (India) using multi-agent systems

Published in Modeling Earth Systems and Environment , 2015

This study explores application of multi agent system (MAS) to simulate spatial patterns of urban growth in Chandigarh and its surrounding region (India). A numerical simulation model is developed with MAS considering the dynamics of urban and rural population as the principal driver of urban growth. The model utilizes static and dynamic environment variables initialized using a logistic regression model. The logistic regression model uses pixel wise change/no-change information derived using Landsat TM data (1989–1999) as dependent variable and proximity, density, elevation and slope as independent variables. The optimum resolution of 90 m for modelling is decided using fractal analysis of series of transition probability surfaces generated using logistic regression from 30 to 240 m spatial resolution at 30 m interval. The model was finally calibrated using sensitivity analysis and behaviours space experiments with multiple simulation runs. A change to built-up area of 32.55 km2 is observed during 1989–1999 and 113.51 km2 in 1999–2009. The modelling shows a total 14.42 % disagreement between predicted map and reference map for the year 2009. The results were validated using ROC statistics and accuracy estimates with satellite data. The model was further used to predict urban growth for the year 2019. Diversity index was used to determine the potential of the model to capture overall spatial patterns of urban growth.

Download here

Urbanization and agricultural land loss in India: Comparing satellite estimates with census data

Published in Journal of Environmental Management , 2015

We examine the impacts of urbanization on agricultural land loss in India from 2001 to 2010. We combined a hierarchical classification approach with econometric time series analysis to reconstruct land-cover change histories using time series MODIS 250 m VI images composited at 16-day intervals and night time lights (NTL) data. We compared estimates of agricultural land loss using satellite data with agricultural census data. Our analysis highlights six key results. First, agricultural land loss is occurring around smaller cities more than around bigger cities. Second, from 2001 to 2010, each state lost less than 1% of its total geographical area due to agriculture to urban expansion. Third, the northeastern states experienced the least amount of agricultural land loss. Fourth, agricultural land loss is largely in states and districts which have a larger number of operational or approved SEZs. Fifth, urban conversion of agricultural land is concentrated in a few districts and states with high rates of economic growth. Sixth, agricultural land loss is predominantly in states with higher agricultural land suitability compared to other states. Although the total area of agricultural land lost to urban expansion has been relatively low, our results show that since 2006, the amount of agricultural land converted has been increasing steadily. Given that the preponderance of India’s urban population growth has yet to occur, the results suggest an increase in the conversion of agricultural land going into the future.

Download here