Google Earth Engine was used to create single date or seasonal
composites of Landsat and NAIP imagery (see below) and compute
statistics for the imagery within PASS boundaries.
Three types of products employing annual remotely sensed images from
Landsat and NAIP: (1) NDVI was computed from the NAIP sensor with a
1-m spatial resolution, and temporal resolution of 2010-2017, (2) SAVI
was computed from the NAIP sensor with a 1-m spatial resolution, and
temporal resolution of 2010-2017, (3) LST was computed from the
Landsat 5 and 8 sensors with a 30-m spatial resolution, and temporal
resolution of 1985-2015.
Landsat and National Agriculture Imagery Program (NAIP) imagery are
used because they are complementary in terms of their spatial and
temporal resolution. Landsat has greater temporal coverage (1985-2015)
but poorer spatial resolution (30m by 30m pixels). NAIP has a more
limited temporal coverage (2010-2017) but high spatial resolution (1m
x 1m pixels).
Tier 1 Surface Reflectance Landsat Imagery was used for calculating
LST. The surface reflectance product has atmospheric correction which
accounts for variations between dates, sensors, and locations (i.e.,
water vapor, ozone, aerosol optical thickness, clouds and digital
elevation) so that the imagery can be used for time-series analysis
(USGS 2018a, 2018b). The NAIP imagery is taken from an airplane, so
while it has a much higher spatial resolution, it may not be reliable
for time-series analysis. NAIP products are best for use in an
analysis that focuses on a single year or for maps/visualizations.
To compute LST from the thermal band of Landsat (band 6 for Landsat 5
and band 10 for Landsat 8), NDVI was used to correct for emissivity
(Shen et al. 2016). Cloudless, summertime (July and August) images
were used for the calculation.
NDVI is computed using the near-infrared (NIR) and red (RED) bands
because red visible light (0.63-0.69 μm) is absorbed by a plant’s
chlorophyll while near-infrared light (0.77-0.90 μm) is scattered by
the leaf’s mesophyll structure: NDVI = (NIR - RED)/(NIR + RED)
SAVI is computed using the same bands as NDVI along with a constant
that corrects for soil brightness (Huete 1988). It is calculated: SAVI
= ((1 + L)(NIR – RED))/(NIR + RED + L) where L = 0.5. SAVI is a
complementary vegetation indice to NDVI in desert regions, such as the
Phoenix metropolitan area, because SAVI minimizes the influence of
soil brightness.
The mean, median, minimum, maximum, and standard deviation of pixel
values were calculated for the each of the neighborhoods in the PASS
(Phoenix Area Social Survey). Neighborhood boundaries vary slightly
over the several years the survey has been conducted—to capture all
variations in boundaries, the statistics were calculated for both the
2011 and 2017 PASS boundaries.
Locations and areas of PASS study neighborhood boundaries are
available through the Environmental Data Initiative:
–PASS 2011:
Harlan S., R. Aggarwal, D. Childers, J. Declet-Barreto, S. Earl, K.
Larson, M. Nation, D. Ruddell, K. Smith, P. Warren, A. Wutich, A.
York. 2018. Phoenix Area Social Survey (PASS): 2011. Environmental
Data Initiative.
https://doi.org/10.6073/pasta/f39a2c9d8e78e6d7a949e93af12e9bf9
–PASS 2017:
Larson K., A. York, R. Andrade, S. Wittlinger. 2019. Phoenix Area
Social Survey (PASS): 2017. Environmental Data Initiative.
https://doi.org/10.6073/pasta/98dd5b92117e9d728b09e582fb4d1b17
References
Huete, A.R., (1988) A soil-adjusted vegetation index (SAVI). Remote
Sensing of Environment, 25 (3): 259-309.
https://doi.org/10.1016/0034-4257(88)90106-X
Shen, Huanfeng, Liwen Huang, Liangpei Zhang, Penghai Wu, and Chao
Zeng. (2016). Long-Term and Fine-Scale Satellite Monitoring of the
Urban Heat Island Effect by the Fusion of Multi-Temporal and
Multi-Sensor Remote Sensed Data: A 26-Year Case Study of the City of
Wuhan in China. Remote Sensing of Environment 172: 109–25.
https://doi.org/10.1016/j.rse.2015.11.005.
U.S. Geological Survey, Department of the Interior. (2018). Landsat
4-7 Surface Reflectance (LEDAPS) Product Guide. LSDS-1370, Version
1.0. EROS: Sioux Falls, South Dakota.
U.S. Geological Survey, Department of the Interior. (2018). Landsat 8
Surface Reflectance (LASRC) Product Guide. LSDS-1368, Version 1.0.
EROS: Sioux Falls, South Dakota.
