2026-08-31

7 arXiv categories· 96 journals· 2053 candidates — 4 worth your time· 98 without an open abstract

Headline

Urban green-patch cooling thresholds, and street-crime modeling with obfuscated data

Today, together

tag shift
no tag ran above its 30-day average
on the same shoulders
no foundation work was cited by four of the day's papers counted over the 96 of this day's 102 named papers whose reference lists we have — standing on the same older work is not the same as being about the same thing
first appearance
Academy of Scientific and Innovative Research; National Botanical Research Institute; University of Lucknow none of them on a paper we published in the 80 days before this issue

Two counts, not one counted twice: every row here but “on the same shoulders” counts the papers published above, and that one counts every paper this day named — those and the ones under Also published today, whose abstracts we could not read.

The first paper analyzed 456 urban and peri-urban green patches in Lucknow, India, using satellite data from 2019 and 2024 to determine how patch size and vegetation condition influence surface-temperature contrasts. Another study developed TransMod, a unified forecasting framework designed to transfer spatio-temporal demand patterns from data-rich transportation modes to data-scarce ones. Finally, researchers evaluated state-of-the-art cross-city point-of-interest recommendation models using the large-scale Trip World benchmark.

published: Discover Cities

Satyabhan Singh, Priya Verma, Soumit K. Behera, et al.
Discover Cities

National Botanical Research Institute

Urban Heat Island Mitigation · Urban Green Space and Health · Land Use and Ecosystem Services · satellite imagery · National Botanical Research Institute · University of Lucknow

Patch size and vegetation condition influence surface-temperature contrasts around urban green patches in a humid subtropical city

The study analyzed 456 urban and peri-urban green patches in Lucknow, India, using Landsat 8 land surface temperature (LST) and Sentinel-2 vegetation indicators from summer and winter scenes in 2019 and 2024. Green patches were cooler than surrounding built-up areas, showing a median warm-season temperature contrast of 0.6–0.8 °C, with cooling effects diminishing for patches larger than 8–10 hectares and flattening beyond a 100–200 meter buffer. Vegetation condition was a stronger predictor of summer cooling than patch shape, and peri-urban patches exhibited stronger temperature contrasts than urban ones.

Why it matters — It establishes specific spatial thresholds for park cooling in a humid subtropical context, demonstrating that cooling benefits plateau beyond 10 hectares and that temperature mitigation is concentrated within 200 meters of the green space boundary.

Caveat: The observed differences between urban and peri-urban patch cooling remain partially confounded by the varying characteristics of their surrounding built-up matrices.

Discover Cities · doi

preprint

Yixuan Zhao, Man Luo
arXiv cs.LG

Traffic Prediction and Management Techniques · Human Mobility and Location-Based Analysis · Transportation and Mobility Innovations

Learning to Transfer Across Modes: Towards Unified Urban Mobility Forecasting

The researchers developed TransMod, a unified forecasting framework that transfers spatio-temporal demand patterns from data-rich transportation modes to data-scarce ones. The model aligns heterogeneous spatial structures into a shared zone-level representation to resolve structural mismatches and distributional shifts across different urban mobility networks.

Why it matters — It enables accurate demand forecasting for newly emerging or data-poor transit modes by leveraging historical data from established, data-rich modes, bypassing the traditional requirement for extensive historical datasets for every individual mode.

Caveat: The performance of the framework is validated on real-world datasets, but the abstract does not specify the number of cities, the specific transit modes tested, or the exact scale of the evaluation data.

arXiv · pdf

preprint

Peibo Li, Yang Song, Hao Xue, et al.
arXiv cs.AI

Human Mobility and Location-Based Analysis · Data Management and Algorithms · Transportation and Mobility Innovations

An Empirical Evaluation of Cross-City POI Recommendation on a Large-Scale Benchmark

This study evaluates state-of-the-art cross-city point-of-interest (POI) recommendation models using the large-scale Trip World benchmark, which features global coverage, low overlap between home and destination regions, and large semantic inventories. The evaluation reveals that existing hometown-aware models rely heavily on destination-region priors rather than transferring user preferences, suffer from severe accuracy-efficiency trade-offs where simple models outperform complex ones, and fail to benefit from semantic metadata. Additionally, a pilot test of LLM-based agentic methods adapted for this task shows they perform worse than a basic popularity baseline.

Why it matters — It exposes critical scalability and design flaws in current cross-city recommendation algorithms, demonstrating that methods optimized on small, localized datasets fail to generalize, transfer preferences, or utilize semantic data when scaled globally.

arXiv · pdf · doi

preprint

Lulu Jiang, David Bolin
arXiv stat.AP

Crime Patterns and Interventions · Crime, Illicit Activities, and Governance · Privacy-Preserving Technologies in Data

Modeling Spatially Obfuscated Street-Crime Data using Log-Gaussian Cox Processes on Metric Graphs

The researchers developed a log-Gaussian Cox process framework to model street-level crime on road networks when the public data has been spatially obfuscated for privacy. Using a Whittle-Matern Gaussian field on a metric graph, the model represents anonymized proxy coordinates as aggregated supports along the street network. Tested on simulated data and real-world crime records from the City of London, the model outperformed planar point and planar aggregated models in parameter recovery and fit across multiple crime types, including theft, robbery, and drug offenses.

Why it matters — It provides a mathematically principled method to analyze network-constrained spatial events using publicly available, anonymized police data without losing the structural constraints of the road network. This allows researchers to accurately assess how environmental amenities, like supermarkets, correlate with crime risk even when exact crime locations are withheld.

Caveat: While the model outperforms planar alternatives, its empirical validation is limited to crime types within a single, highly specific urban jurisdiction, the City of London.

arXiv · pdf

Still cited

Diana E. Bowler, Lisette M Buyung-Ali (2010), Urban greening to cool towns and cities: A systematic review of the empirical evidence 1 of today's items cite it · 30 of 3240 in the archive stand on it

Also published today

These appeared today in journals we track. Their abstracts are not openly available, so we cannot summarise them.