https://doi.org/10.5061/dryad.rbnzs7hj6

Summary

This dataset includes estimates of temperature’s effects on dengue that were collected from the literature; however, only a subset of this database was used in the analyses for this study due to reasons described in the methods (e.g., correlations were preferred to regression coefficients). Accompanying R code illustrates which data is used in each analysis.
 
Below is a short description of the data features, not all of which were used in the analyses conducted in the study. More explanation for how these data were collected or calculated is included in the manuscript and supporting information.
 
study_code: Factor to define different studies in the literature review.
reference: First author name and study year.
study_location: Study location.
study_scale: Broad scale of the study (e.g., city, country).
long: Longitude the study occurred at. NA if the study scale was large enough that shapefiles were used to define where the study occurred.
lat: latitude the study occurred at. NA if the study scale was large enough that shapefiles were used to define where the study occurred.
unique_ID: Factor for identifying studies.
continent: Continent where the study occurred.
country_or_territory: Country or territory where the study occurred.
study_length_in_days: Length of the study in days.
start_date: Day the study began.
end_date: Day the study ended.
specific_temperature_metric: Reported temperature metric used to estimate the effect of temperature on dengue.
broad_temperatre_metric: Temperature metric used to estimate the effect of temperature on dengue categorized as “mean”, “min”, or “max”.
specific_disease_metric: Reported dengue disease metric used when estimating the effect of temperature on dengue.
broad_disease_metric: Dengue disease metric used when estimating the effect of temperature on dengue categorized as “cases” or “incidence”.
data_temporal_scale: Temporal scale of the data (e.g. “weekly”, “monthly”).
average_lag_in_months: Average lag in temperature (unit = months) used when estimating temperature effect on dengue.
model_used: Type of model used to estimate temperature effect on dengue (e.g. “Correlation”, “Poisson GLM”).
sample_size: Number of temperature–dengue datapoints used in model.
number_covariates: Number of covariates included in model.
latitude_weighted: Population-weighted latitude where the study occurred.
mean_2m_air_temperature_mean_weighted: Population-weighted average temperature where the study occurred, extracted from Google Earth Engine. Unit is Kelvin.
mean_2m_air_temperature_stDev_weighted: Population-weighted standard deviation in temperature where the study occurred, extracted from Google Earth Engine.
pop_per_m2_unweighted: Population per m2 where the study occurred.
pop_per_m2_weighted: Population-weighted population per m2 where the study occurred.
population_year: Year that population size was estimated from.
total_precipitation_mean_weighted: Population-weighted precipitation where the study occurred, extracted from Google Earth Engine. Precipitation measured as daily sum in meters.
total_precipitation_stdDev_weighted: Population-weighted precipitation where the study occurred, extracted from Google Earth Engine.
inapparent_infection_burden_2010: Dengue infection burden for 2010 estimated by Bhatt et al. (2013).
population_size_2010: Population size in 2010.
inapparent_infection_incidence_2010: Inapparent infection incidence in 2010 per 100,000 people.
per_capita_GDP_PPP_2015: Per capita GDP adjusted for purchasing price parity in 2015.
mean_temp_C: Mean_2m_air_temperature_mean_weighted converted from Kelvin into degrees C.
log_pop_per_m2_weighted: Natural log of pop_per_m2_weighted.
effect_size: Effect of temperature on dengue as estimated by the model.
lower_95_CI: Lower 95% confidence interval value around estimated effect_size. Often not reported, especially if model being used is a correlation.
upper_95_CI: Upper 95% confidence interval value around estimated effect_size. Often not reported, especially if model being used is a correlation.
effect_type: Type of effect associated with the reported effect_size (e.g., “RR” for relative risk, “spearman” for spearman correlation”.
coef_or_cor: Effect_type categorized as either a regression coefficient or a correlation.
exclude_cross_correlations: TRUE if the effect_type is not a cross-correlation, FALSE if the effect_type is a cross-correlation. Can be used to compare cross-correlation effects versus non-cross-correlation effects if desired.
temperature_specific_analysis: Binary 0 or 1 factor used to subset the data used in the temperature-specific analysis compared to the analysis investigating other socioeconomic factors.

In the "dengueR0.csv" file, "temp" represents temperature in degrees C, and "R0" is the predicted dengue basic reprodution number at that temperature based on the models in Mordecai et al. (2017, PLOS Negl. Trop. Dis.)

R Code

Affiliated R code is available at https://github.com/devingkirk.