Mr Lei Wang (p), Prof. Kaichun Zhou, Prof. Chunla Liu, Prof. Binggeng Xie, Prof. Xiaoqing Li, Prof. Jiancheng Zheng

In recent years, the conflict between ensuring food security, maintaining ecological functions, and promoting industrial upgrading has become increasingly prominent in the Changsha-Zhuzhou-Xiangtan (CZT) metropolitan area. Based on land-use data from 2015 to 2023, this study utilizes spatial analysis of land-use transitions and Geographical Detectors to analyze the evolution of land-use patterns in terms of structure, quantity, and spatial morphology, while exploring the underlying driving mechanisms. The results indicate that: (1) The overall land-use structure of the CZT metropolitan area maintains a fundamental pattern of "ecological dominance, agricultural foundation, and rapid expansion of construction land." (2) Ecological and agricultural spaces remain dominant, forming the essential base for regional ecological security and food production, while the proportion of construction land continues to rise driven by urbanization. (3) Land use exhibits a typical "two-decrease, three-increase, one-stable" pattern: cropland and forest land have overall declined; construction land, water bodies, and unused land have grown steadily; and grassland has remained largely stable. (4) The expansion trajectory of construction land has evolved from early "point-like growth" to "block-like clustering," gradually forming a "belt-like continuous" structure centered on main transportation axes and the boundary areas of the three cities, with land elements continuously converging toward the metropolitan core and axial zones. (5) The evolutionary dynamics consist of external pushes from national strategies and policy systems, internal pulls from market mechanisms, industrial clusters, and population mobility, as well as the constraints and guidance provided by ecological red lines, "Green Heart" management, and spatial structural evolution.

Ms Ragdad Cani Miranti (p), Mr David Lawson

Night-time light (NTL) has emerged as an increasingly prominent source in socioeconomic research over the past decades, widely recognised as a reliable proxy for regional economic activity. Across developing and emerging economies, official statistics, such as, Gross Regional Domestic Product can be prone to time delays in dissemination. Using NTL captured by satellite sensors, we provide spatial and time sectoral output dynamics, across a decade to 2024, examining the interplay between urban-rural luminosity and sectoral output in agriculture, industry, and services in Indonesia. We apply Visible Infrared Imaging Radiometer Suite (VIIRS) NTL which offers advantages on higher spatial resolution and finer detail of geographical areas compared to previous Defense Meteorological Satellite Program (DMSP) data. To decompose night-time light into urban-rural specification, we apply Moderate Resolution Spectroradiometer Land Cover data (MODIS/061/MCD12Q1) and overlay with Global Human Settlement Layer from Google Earth Engine. We also apply GRDP data across 514 districts of Indonesia, over the 2014-2024 period, incorporated with other remote-sensing driven, population, and investment variables to conceptualise the framework of Solow (1956) and Romer (1986) model. Our empirical strategy employs four robustness checks: between and within regression approaches to assess cross-sectional and temporal variation; comparative analysis across pre-COVID-19 (2014-2019), during COVID-19 (2020-2021), and full period; compare predictive power and fit both at aggregate and growth level; and exploratory spatio-temporal analysis examining spatial patterns and co-evolution across Indonesia's eight major islands using Rose Diagram and Directional LISA.
Results reveal three key findings. First, NTL data demonstrates significant predictive power for both cross-sectional and temporal economic variation, with coefficients remaining robust across different periods and after controlling for socioeconomic variables, except during COVID-19 period. Specifically, elasticities of NTL to predict sectoral economies is higher during pre-COVID 19 period compared to COVID-19 period. Second, urban NTL exhibits better predictive fit for industry and services output, while rural NTL better captures agricultural output. Spatial disaggregation across full periods (2014-2024) reveals pronounced sectoral heterogeneity in NTL elasticities, in which 10 percent increase in urban NTL corresponds to output increases of 6.04 percent in the industrial sector and 3.21 percent in the service sector, reflecting the urban concentration of high-productivity economic activities. Third, spatio-temporal analysis demonstrates positive co-evolution patterns between sectoral output and urban light, with Java-Bali and Sumatra showing the highest frequency, while Kalimantan and Papua exhibit the highest frequency on rural patterns. These findings demonstrate convergence in regional economic dynamics and reduced urban-rural spatial polarisation.

Prof. Piotr Wójcik (p), Patrick Sliz

This paper examines whether nighttime lights imagery (NTLI) can reliably explain and predict regional economic well‑being in a highly developed country using only remotely sensed data. Building on prior work that has largely focused on cross‑sections, developing economies, single predictors, and short time spans, we construct a multi‑year, multi‑scale framework for the United States using the Harmonized Global Nighttime Lights series, which combines DMSP‑OLS and VIIRS into a consistent 1992–2022 panel. We derive a rich set of NTLI‑based indicators at state and county level, capturing not only total brightness but also distributional and spatial characteristics of lights within each area.

We link these features to official data on real GDP and personal income, and estimate models separately for explanatory (same‑year) fit and predictive performance (one‑year‑ahead forecasts). Methodologically, we compare a baseline log‑log OLS specification with several non‑linear machine‑learning algorithms, including LASSO, Support Vector Regression, k‑Nearest Neighbours, Random Forest, and XGBoost. Model performance is assessed using out‑of‑sample R2, and we employ permutation‑based feature importance to identify which NTLI‑derived predictors contribute most to explanatory and predictive accuracy.

The results confirm that the NTLI-economy relationship is substantially stronger at the regional (state) than at the local (county) level, both for GDP and income. Augmenting the sum of lights with additional distributional and spatial metrics yields sizeable gains in explanatory power, especially for states, indicating that how light is distributed across space carries information beyond aggregate brightness. Non‑linear ML models, particularly tree‑based methods, systematically outperform OLS in both explanatory and nowcasting exercises, with the largest improvements again observed at the state level. Our results suggest that harmonized nighttime lights can serve as a practical early indicator of regional economic conditions in advanced economies.

Mrs Maria Adamakou, Prof. Dimitris Kallioras (p)

The paper discusses whether – and to what extent – sub-national population change responds positively or negatively to agglomeration dynamics (i.e., population density), in ways that reinforce or, instead, equalize population concentrations across space. Examining processes of concentration and de-concentration of population goes beyond the demographic interest per se, as it may, also, shed light on questions concerning economic convergence and divergence in the more general sense of understanding wider economic processes. To this end, the paper presents evidence from China. Under the conditions of the rapid market liberalization process that China has been experiencing, questions of spatial cohesion – and thus of convergence and divergence – even though are still, rather, neglected, become increasingly salient. This is so as the elimination of spatial imbalances is both a pre-condition and a core objective of the reforms aiming at market liberalization. Scholars both in the urban economics and the growth economics tradition have well-recognized that studying population growth offers a window through which to study the process of economic growth as the latter is systematically related to population growth through the trade-off between agglomeration economies and urban costs (i.e., commuting, housing, land use, environmental, inter alia). This is especially so in China given that productivity levels, capital deepening and levels of technology are rather low, compared to the corresponding levels of the EU and the USA), and thus economic growth is still very much of the extensive (i.e. increase in inputs, including labor) than the intensive form (i.e. increase in the productivity of each input and in total factor productivity). Moreover, given that China is still at low levels of economic development, spatial productivity differentials are predominantly along the lines of urban-rural (i.e. core–periphery) dichotomy, and thus very much related to patterns of population agglomeration (i.e. urbanization). In these conditions, measures of spatial disparity in population concentration act as a lower bound indicator of spatial disparities in terms of economic development, and this may be of acute analytical and policy interest. If patterns of population growth are found to be cumulative across space, this could be taken as a signal of a wider spatial disequilibrium, representing spatial inequalities in economic opportunities more generally. This may, ultimately, raise concerns as regards the success of the market liberalization process in itself.