Supplementary Materials for
Unequal exposure to heatwaves in Los Angeles: Impact of uneven
green spaces
Yi Yin
et al.
Corresponding author: Yi Yin, yinyi11@gmail.com
Sci. Adv.
9
, eade8501 (2023)
DOI: 10.1126/sciadv.ade8501
This PDF file includes:
Figs. S1 to S6
Figure
S1. Spatial distribution of
median household income, ECOSTRESS
Land Surface
Temperature (LST)
, and Landsat
Normalized Difference Vegetation Index (NDVI)
across
Los Angeles County. (a)
Median household income at block group level
,
(b)
LST at 70 m
resolution,
(c)
Landsat NDVI
at 30 m resolution
. The data displayed here at their original
resolutions support the zip code level patterns shown in Fig. 1.
Figure S2. Scatter plot between median household income and ECOSTRESS Land Surface
Temperature (LST) at a finer spatial
resolution
. Scatter plot shown at
(a)
block group level and
(b)
at the LST pixel level, with corresponding income data at the block
-
group level. The spatial
distributions of these variables are shown in Fig
ure
S1
.
Fig
ure
S3
.
Correlations among evapot
ranspiration
(ET)
, land surface temperature
(LST),
and
Normalized Difference Vegetation Index
(
NDVI
)
across the Los Angeles County.
(a)
Correlation between ET
and LST derived from ECOSTRESS
.
(b)
Correlation between the
Landsat
NDVI and instantaneous
ET
from ECOSTRESS. Each dot represents a ZIP Code
Tabulation Area.
ECOSTRESS observations were made at 1 pm on June 4th, 2021, while
Landsat NDVI observations represent monthly averages from June 2021.
Figure S4. Diurnal
distribution slopes of Land Surface Temperature (LST) against albedo,
and correlation coefficients between LST and albedo.
Note that a higher albedo would result
in more reflected radiation and less absorbed energy, leading to negative correlations between
albedo and LST if albedo is the dominant factor impacting the surface energy budget.
The positive
correlations here suggest that albedo is not the dominating factor in shaping the spatial
differences in LST, but rather through the effect of ET.
Fi
g
ure
S5
.
Scatter plot between median household income and elevation, median slope
aspects, and distance to the nearest coast
. At the zip
-
code level, there does not appear to be
a significant correlation between income and topographical features (elevation and slope aspects).
The slope aspect represents the orientation of slope, where 0 is north
-
facing, 90 is east
-
facing,
180 is
south
-
facing, and 270 is west
-
facing. At this scale, the slope orientation is mostly influenced
by the overall landscape other than local topography.
We find that
distance to the nearest coast
is weakly correlated with median household income (r = 0.1, p<
0.01), and this distance could
explain ~10% of the observed gradient in LST (R
2
=0.1, p<0.01).
Fig
ure
S6
.
Analysis of gridded air temperature at 4 km resolution from the PRISM datasets.
(a) Scatter plot of daily maximum air temperature from the PRI
SM datasets and collocated LST
from ECOSTRESS observations at 1 pm, June 4
th
, 2021
(r=0.55, p<0.01)
.
(b) Scatter plot of
median household income (x
-
axis) and PRISM monthly mean values of daily maximum
temperature for the summer months in 2021 (y
-
axis). The
slopes in Fig. S
6
b are negative, but not
statistically significant (p>0.1).