Table 1. Results of univariate logistic regression analysis (frequentist) of candidate covariates evaluated in the study (P<0.2).
Source/variable | Control (Mean ± S.E) | Case (Mean ± S.E.) | p ≤ 0.2 |
---|---|---|---|
National Land Cover Dataset | |||
a. Percentage ground cover | |||
Open water | 0.83 ± 0.18 | 0.75 ± 0.11 | 0.65 |
Developed—open space | 1.54 ± 0.54 | 1.08 ± 0.81 | 0.41 |
Developed—low intensity | 1.41 ± 1.01 | 3.02 ± 0.77 | 0.02 |
Developed—high intensity | 2.44 ± 1.63 | 2.11 ± 2.16 | 0.42 |
Barren land | 1.81 ± 0.68 | 2.21 ± 1.01 | 0.54 |
Deciduous forest | 8.41 ± 2.74 | 9.27 ± 2.33 | 1.48 |
Mixed forest | 2.14 ± 1.20 | 2.98 ± 0.58 | 0.29 |
Evergreen forest | 10.24 ± 3.54 | 8.25 ± 2.89 | 0.61 |
Scrub/shrub | 7.24 ± 2.39 | 6.74 ± 3.04 | 0.74 |
Grassland/herbaceous cover | 0.84 ± 0.01 | 2.11 ± 0.20 | 0.09 |
Pasture/hay | 12.36 ± 5.63 | 10.51 ± 5.21 | 0.35 |
Woody wetlands | 1.36 ± 0.85 | 2.11 ± 1.02 | 0.71 |
Emergent herbaceous wetlands | 1.55 ± 1.02 | 1.34 ± 0.87 | 1.88 |
b. Landscape metric | |||
Total edge contrast index (TECI) | 1.84 ± 0.21 | 2.81 ± 0.42 | 0.03 |
NASA Moderate Resolution Imaging Spectrometer (MODIS) | |||
a. Daytime land surface temperature | |||
≥35°C | 36.32 ± 3.14 | 38.21 ± 2.36 | 0.61 |
28–34.9°C | 30.95 ± 1.25 | 29.22 ± 1.23 | 1.10 |
24.9–27.9°C | 26.54 ± 1.02 | 26.11 ± 0.84 | 1.51 |
≤25°C | Reference category | ||
b. Night time land surface temperature | |||
≤16°C | Reference category | ||
15.9–19.9°C | 16.44 ± 1.21 | 16.14 ± 1.25 | 1.21 |
≥20°C | 21.33 ± 0.88 | 23.01 ± 1.05 | 0.19 |
c. Diurnal temperature range | 12.00 ± 0.14 | 15.21 ± 0.41 | 0.04 |
NASA Prediction of Worldwide Renewable Resources | |||
Daily maximum temperature | 34.32 ± 1.22 | 34.62 ± 1.24 | 1.84 |
Daily minimum temperature | 15.25 ± 1.55 | 16.05 ± 1.50 | 0.57 |
Daily average temperature | 28.44 ± 2.61 | 27.58 ± 2.07 | 0.66 |
Dew point | 63.32 ± 10.21 | 59.21 ± 7.32 | 0.75 |
Relative humidity | 74.21 ± 8.21 | 75.38 ± 6.32 | 0.81 |
Diurnal temperature range | 11.14 ± 1.51 | 13.21 ± 1.32 | 0.06 |
A total of 23 variables were considered for univariate evaluations. There were 14 variables derived from the NLCD, 3 variables from NASA’s MODIS and 6 from POWER sources. All variables except daytime land surface temperature and night-time land surface temperature were in continuous form. ≤25°C and ≤16°C were used as reference categories in the models for daytime land surface temperature and night-time land surface temperature, respectively. Six variables retained significance in the univariate screening, with a liberal p−value ≤ 0.2. They were, developed—low intensity (p = 0.021), grassland/herbaceous cover (p = 0.092), grassland/herbaceous cover (p = 0.122), total edge contrast index (p = 0.033), night time land surface temperature (≥20°C) (p = 0.191), and diurnal temperature range (DTR) (p = 0.044).