Panel Hot Spot Analysis
Panel Data Hot Spot Analysis · Also known as: longitudinal hot spot analysis, repeated cross-sectional hot spot analysis, spatio-temporal hot spot detection, panel Getis-Ord analysis
Panel Hot Spot Analysis applies hot spot detection — typically via the Getis-Ord Gi* statistic — repeatedly across multiple time periods on the same spatial units, enabling researchers to track where clusters of high or low values persist, emerge, or dissolve over time. It bridges cross-sectional spatial statistics with longitudinal panel methods.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use Panel Hot Spot Analysis when you have georeferenced data measured repeatedly over time on the same spatial units and your goal is to characterise the temporal stability or change of spatial clusters. It is well-suited for crime analysis, public health surveillance, environmental monitoring, and urban studies where understanding whether patterns are chronic or transient matters. Do not use it when data are available for only one time period (use standard hot spot analysis instead), when spatial units change across periods making comparison invalid, or when the temporal resolution is so coarse that within-period dynamics are lost.
Strengths & limitations
- Captures temporal dynamics of spatial clusters — distinguishing chronic, emerging, and dissipating hot spots.
- Leverages the well-validated Gi* statistic, producing z-scores with clear statistical interpretation.
- Provides a rich basis for policy targeting: chronic hot spots warrant different interventions than emerging ones.
- Compatible with group-based trajectory modelling and other longitudinal analytic tools.
- Applicable in diverse domains: criminology, epidemiology, ecology, and urban planning.
- Requires spatially consistent units across all time periods; boundary changes (e.g., redistricting) break comparability.
- The spatial weights matrix is typically held fixed, which may not capture evolving neighbourhood relationships.
- Does not account for spatial autocorrelation in the panel residuals by itself; supplementary spatial panel models are needed for covariate inference.
- Multiple testing across many units and periods inflates Type I error unless FDR or similar corrections are applied.
- Computationally intensive for large panels with many spatial units and long time series.
Frequently asked
How is Panel Hot Spot Analysis different from standard hot spot analysis?
Standard hot spot analysis identifies clusters at a single point in time. Panel Hot Spot Analysis computes the Gi* statistic at multiple time periods and examines how each spatial unit's hot-spot status evolves, enabling detection of chronic, emerging, or dissipating clusters.
What spatial weights matrix should I use?
Use a fixed weights matrix (e.g., queen contiguity or a distance band defined once) applied consistently across all periods. This ensures that Gi* z-scores are comparable over time. Row-standardise the matrix for interpretability.
How do I classify hot-spot trajectories?
Common approaches include: (1) labelling a unit a 'chronic hot spot' if its Gi* z-score exceeds a threshold in a majority of periods; (2) ArcGIS's Emerging Hot Spot Analysis algorithm, which uses the Mann-Kendall trend test on the Gi* time series; and (3) group-based trajectory modelling to fit distinct trajectory groups empirically.
How do I handle multiple comparisons?
With many spatial units and time periods, numerous simultaneous tests are conducted. Apply a false discovery rate (FDR/Benjamini-Hochberg) correction or, more conservatively, Bonferroni correction to the p-values of the Gi* statistics before classifying significance.
Can I add covariates to explain why certain units are persistent hot spots?
Yes. After deriving trajectory classifications or time-series Gi* scores per unit, use panel regression models (fixed or random effects) with the Gi* score or hot-spot indicator as the outcome and covariates of interest as predictors. Spatial lag or spatial error panel models should be considered to account for residual spatial dependence.
Sources
- Weisburd, D., Bushway, S., Lum, C., & Yang, S.-M. (2004). Trajectories of crime at places: A longitudinal study of street segments in the city of Seattle. Criminology, 42(2), 283-321. DOI: 10.1111/j.1745-9125.2004.tb00521.x ↗
- Getis, A., & Ord, J. K. (1992). The analysis of spatial association by use of distance statistics. Geographical Analysis, 24(3), 189-206. DOI: 10.1111/j.1538-4632.1992.tb00261.x ↗
How to cite this page
ScholarGate. (2026, June 3). Panel Data Hot Spot Analysis. ScholarGate. https://scholargate.app/en/spatial-analysis/panel-hot-spot-analysis
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Hot Spot AnalysisSpatial analysis↔ compare
- Local Getis-Ord Gi*Spatial analysis↔ compare
- Local Indicators of Spatial AssociationSpatial analysis↔ compare
- Space-Time Hot Spot AnalysisSpatial analysis↔ compare