Urban Heat (Temperature), NDVI, and Cluster Analysis

Introduction

Temperature and NDVI are two major factors when considering Urban Heat Index Selection of Landsat 7 Data for Consistency with Census Intervals The Urban Heat Index (UHI) analysis in Hong Kong heavily relies on two primary factors: temperature and the Normalized Difference Vegetation Index (NDVI). To align the satellite data with census intervals, we’ve opted for data from Landsat 7 ETM+. The choice of Landsat 7, despite the availability of Landsat 8 and 9, is dictated by the need for consistency with the census data from the Hong Kong government and ESRI Hong Kong, which are available every five years. The most recent census data was from 2021, necessitating the use of data from 2011 and 2016 as well. Given that Landsat 8 started operating in 2013, it wouldn’t cover the entire period corresponding to our census data, leading to the decision to utilize Landsat 7’s data for a comprehensive analysis.

Methodology for Accurate Data Retrieval and Processing

For ensuring accuracy, our approach involved creating a mosaic composite raster image from Google Earth Engine, filtered for the entire year. This process was repeated for each year from 2011 to 2021 to extract the annual mean of temperature and NDVI for each pixel. The most challenging aspect of this section was the calibration of the data retrieved from Google Earth Engine. We encountered numerous difficulties in calibrating the band data. For instance, there were times when the NDVI worked correct (in the range) on Google Earth Engine but failed in Python. Similarly, issues were faced with the land surface temperature data. After countless attempts and explorations, we discovered a calculation method from an academic paper that provided a more refined approach. This method allowed for the conversion of digital numbers, surface reflectance, top of the atmosphere reflectance, top brightness temperature, and land surface temperature in a more accurate way.

(GEE link:https://code.earthengine.google.com/a1c8085a46f0c89f1d82f0cfcfabacd8)

Reference: Ermida, S.L., Soares, P., Mantas, V., Göttsche, F.-M., Trigo, I.F., 2020. Google Earth Engine open-source code for Land Surface Temperature estimation from the Landsat series. Remote Sensing, 12 (9), 1471; https://doi.org/10.3390/rs12091471

Load the Lansat 7 2011 Mosaic Image

<xarray.Dataset>
Dimensions:      (band: 16, x: 2503, y: 1783)
Coordinates:
  * band         (band) int64 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
  * x            (x) float64 113.8 113.8 113.8 113.8 ... 114.5 114.5 114.5 114.5
  * y            (y) float64 22.57 22.57 22.57 22.57 ... 22.09 22.09 22.09 22.09
    spatial_ref  int64 ...
Data variables:
    band_data    (band, y, x) float32 ...

xarray.Dataset

• Dimensions:
  • band: 16
  • x: 2503
  • y: 1783
• Coordinates: (4)
  • band
    (band)
    int64
    1 2 3 4 5 6 7 ... 11 12 13 14 15 16

    array([ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16])

  • x
    (x)
    float64
    113.8 113.8 113.8 ... 114.5 114.5

    array([113.784791, 113.785061, 113.78533 , ..., 114.458528, 114.458797,
           114.459067])

  • y
    (y)
    float64
    22.57 22.57 22.57 ... 22.09 22.09

    array([22.57354 , 22.573271, 22.573001, ..., 22.09384 , 22.09357 , 22.093301])

  • spatial_ref
    ()
    int64
    ...

    crs_wkt :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    semi_major_axis :

    6378137.0

    semi_minor_axis :

    6356752.314245179

    inverse_flattening :

    298.257223563

    reference_ellipsoid_name :

    WGS 84

    longitude_of_prime_meridian :

    0.0

    prime_meridian_name :

    Greenwich

    geographic_crs_name :

    WGS 84

    horizontal_datum_name :

    World Geodetic System 1984

    grid_mapping_name :

    latitude_longitude

    spatial_ref :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    GeoTransform :

    113.78465630535715 0.00026949458523585647 0.0 22.573674943111044 0.0 -0.00026949458523585647

    [1 values with dtype=int64]

• Data variables: (1)
  • band_data
    (band, y, x)
    float32
    ...

    AREA_OR_POINT :

    Area

    long_name :

    ('SR_B1', 'SR_B2', 'SR_B3', 'SR_B4', 'SR_B5', 'SR_B7', 'QA_PIXEL', 'NDVI', 'FVC', 'TPW', 'TPWpos', 'EM', 'B6_VCID_1', 'B6_VCID_2', 'LST', 'BBE')

    [71405584 values with dtype=float32]

• Indexes: (3)
  • band
    PandasIndex

    PandasIndex(Int64Index([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16], dtype='int64', name='band'))

  • x
    PandasIndex

    PandasIndex(Float64Index([113.78479105264977,   113.785060547235, 113.78533004182023,
                  113.78559953640547, 113.78586903099071, 113.78613852557595,
                  113.78640802016118, 113.78667751474642, 113.78694700933166,
                   113.7872165039169,
                  ...
                  114.45664105364276,   114.456910548228, 114.45718004281322,
                  114.45744953739846,  114.4577190319837, 114.45798852656894,
                  114.45825802115417,  114.4585275157394, 114.45879701032464,
                  114.45906650490988],
                 dtype='float64', name='x', length=2503))

  • y
    PandasIndex

    PandasIndex(Float64Index([22.573540195818428,  22.57327070123319, 22.573001206647955,
                   22.57273171206272, 22.572462217477483, 22.572192722892247,
                  22.571923228307014, 22.571653733721778, 22.571384239136542,
                  22.571114744551306,
                  ...
                  22.095726296195256,  22.09545680161002, 22.095187307024784,
                  22.094917812439547,  22.09464831785431, 22.094378823269075,
                   22.09410932868384, 22.093839834098603, 22.093570339513366,
                   22.09330084492813],
                 dtype='float64', name='y', length=1783))

• Attributes: (0)

Read the Hong Kong District Data from ArcGIS geojson

geopandas.geodataframe.GeoDataFrame

We decided to use EPSG: 4326 since the CRS of Landsat 7 Data’s is EPSG= 4326

(113.835066606914, 22.153344109236, 114.441993257571, 22.5619493489016)

Trim our precious Raster Data to the Hong Kong Boundary

<xarray.Dataset>
Dimensions:      (band: 16, x: 2252, y: 1516)
Coordinates:
  * band         (band) int64 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
  * x            (x) float64 113.8 113.8 113.8 113.8 ... 114.4 114.4 114.4 114.4
  * y            (y) float64 22.56 22.56 22.56 22.56 ... 22.15 22.15 22.15 22.15
    spatial_ref  int64 0
Data variables:
    band_data    (band, y, x) float32 1.004e+04 9.852e+03 ... 0.9903 0.9901

xarray.Dataset

• Dimensions:
  • band: 16
  • x: 2252
  • y: 1516
• Coordinates: (4)
  • band
    (band)
    int64
    1 2 3 4 5 6 7 ... 11 12 13 14 15 16

    array([ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16])

  • x
    (x)
    float64
    113.8 113.8 113.8 ... 114.4 114.4

    array([113.835187, 113.835456, 113.835726, ..., 114.44128 , 114.441549,
           114.441819])

  • y
    (y)
    float64
    22.56 22.56 22.56 ... 22.15 22.15

    array([22.561682, 22.561413, 22.561143, ..., 22.153937, 22.153668, 22.153398])

  • spatial_ref
    ()
    int64
    0

    crs_wkt :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    semi_major_axis :

    6378137.0

    semi_minor_axis :

    6356752.314245179

    inverse_flattening :

    298.257223563

    reference_ellipsoid_name :

    WGS 84

    longitude_of_prime_meridian :

    0.0

    prime_meridian_name :

    Greenwich

    geographic_crs_name :

    WGS 84

    horizontal_datum_name :

    World Geodetic System 1984

    grid_mapping_name :

    latitude_longitude

    spatial_ref :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    GeoTransform :

    113.78465630535715 0.00026949458523585647 0.0 22.573674943111044 0.0 -0.00026949458523585647

    array(0)

• Data variables: (1)
  • band_data
    (band, y, x)
    float32
    1.004e+04 9.852e+03 ... 0.9901

    AREA_OR_POINT :

    Area

    long_name :

    ('SR_B1', 'SR_B2', 'SR_B3', 'SR_B4', 'SR_B5', 'SR_B7', 'QA_PIXEL', 'NDVI', 'FVC', 'TPW', 'TPWpos', 'EM', 'B6_VCID_1', 'B6_VCID_2', 'LST', 'BBE')

    array([[[1.0039000e+04, 9.8520000e+03, 9.8520000e+03, ...,
             9.3790000e+03, 9.5540000e+03, 9.2050000e+03],
            [9.9450000e+03, 1.0039000e+04, 1.0039000e+04, ...,
             9.2910000e+03, 9.2050000e+03, 9.3790000e+03],
            [9.8520000e+03, 9.9460000e+03, 9.9460000e+03, ...,
             9.4680000e+03, 9.2910000e+03, 9.2050000e+03],
            ...,
            [1.0704000e+04, 1.0804000e+04, 1.0804000e+04, ...,
             9.5450000e+03, 9.7170000e+03, 9.8020000e+03],
            [1.0804000e+04, 1.0906000e+04, 1.0604000e+04, ...,
             9.6320000e+03, 9.8020000e+03, 9.7170000e+03],
            [1.0804000e+04, 1.0604000e+04, 1.0906000e+04, ...,
             9.5450000e+03, 9.5450000e+03, 9.6320000e+03]],
    
           [[1.0514000e+04, 1.0614000e+04, 1.0614000e+04, ...,
             9.1010000e+03, 9.1010000e+03, 9.0070000e+03],
            [1.0614000e+04, 1.0514000e+04, 1.0514000e+04, ...,
             9.1940000e+03, 9.1010000e+03, 9.1940000e+03],
            [1.0414000e+04, 1.0514000e+04, 1.0614000e+04, ...,
             9.2880000e+03, 9.1010000e+03, 9.1940000e+03],
    ...
            [2.9195679e+02, 2.9195679e+02, 2.9195679e+02, ...,
             2.9790427e+02, 2.9790427e+02, 2.9790427e+02],
            [2.9195679e+02, 2.9195679e+02, 2.9195679e+02, ...,
             2.9790427e+02, 2.9790427e+02, 2.9790427e+02],
            [2.9137042e+02, 2.9137042e+02, 2.9195679e+02, ...,
             2.9790427e+02, 2.9790427e+02, 2.9790427e+02]],
    
           [[9.8962820e-01, 9.8978066e-01, 9.8978066e-01, ...,
             9.8911905e-01, 9.8882079e-01, 9.8896593e-01],
            [9.8954946e-01, 9.9022216e-01, 9.9022216e-01, ...,
             9.8911905e-01, 9.8875976e-01, 9.8911905e-01],
            [9.9013770e-01, 9.8986351e-01, 9.8978066e-01, ...,
             9.8896772e-01, 9.8926890e-01, 9.8865867e-01],
            ...,
            [9.9023283e-01, 9.8983717e-01, 9.9028999e-01, ...,
             9.9008703e-01, 9.9028999e-01, 9.9008703e-01],
            [9.9023283e-01, 9.8974222e-01, 9.9004382e-01, ...,
             9.9028999e-01, 9.9008703e-01, 9.9028999e-01],
            [9.8992366e-01, 9.8974222e-01, 9.9023283e-01, ...,
             9.9028999e-01, 9.9028999e-01, 9.9008703e-01]]], dtype=float32)

• Indexes: (3)
  • band
    PandasIndex

    PandasIndex(Int64Index([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16], dtype='int64', name='band'))

  • x
    PandasIndex

    PandasIndex(Float64Index([113.83518654008887, 113.83545603467411, 113.83572552925935,
                  113.83599502384457, 113.83626451842981, 113.83653401301505,
                  113.83680350760028, 113.83707300218552, 113.83734249677076,
                    113.837611991356,
                  ...
                  114.43939340018765,  114.4396628947729, 114.43993238935813,
                  114.44020188394337,  114.4404713785286, 114.44074087311384,
                  114.44101036769908, 114.44127986228432, 114.44154935686954,
                  114.44181885145478],
                 dtype='float64', name='x', length=2252))

  • y
    PandasIndex

    PandasIndex(Float64Index([ 22.56168243406805, 22.561412939482814, 22.561143444897578,
                   22.56087395031234, 22.560604455727105,  22.56033496114187,
                  22.560065466556633,   22.5597959719714, 22.559526477386164,
                  22.559256982800928,
                  ...
                   22.15582358870285, 22.155554094117615,  22.15528459953238,
                  22.155015104947143, 22.154745610361907,  22.15447611577667,
                  22.154206621191435,   22.1539371266062, 22.153667632020962,
                  22.153398137435726],
                 dtype='float64', name='y', length=1516))

• Attributes: (0)

Raster image combined with City limits

Calculate the NDVI using Near Infrared Band and Red Band.

<xarray.Dataset>
Dimensions:      (y: 1516, x: 2252)
Coordinates:
  * x            (x) float64 113.8 113.8 113.8 113.8 ... 114.4 114.4 114.4 114.4
  * y            (y) float64 22.56 22.56 22.56 22.56 ... 22.15 22.15 22.15 22.15
    spatial_ref  int64 0
Data variables:
    band_data    (y, x) float32 -0.0636 -0.06632 -0.06632 ... -0.03526 -0.02509

xarray.Dataset

• Dimensions:
  • y: 1516
  • x: 2252
• Coordinates: (3)
  • x
    (x)
    float64
    113.8 113.8 113.8 ... 114.4 114.4

    array([113.835187, 113.835456, 113.835726, ..., 114.44128 , 114.441549,
           114.441819])

  • y
    (y)
    float64
    22.56 22.56 22.56 ... 22.15 22.15

    array([22.561682, 22.561413, 22.561143, ..., 22.153937, 22.153668, 22.153398])

  • spatial_ref
    ()
    int64
    0

    crs_wkt :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    semi_major_axis :

    6378137.0

    semi_minor_axis :

    6356752.314245179

    inverse_flattening :

    298.257223563

    reference_ellipsoid_name :

    WGS 84

    longitude_of_prime_meridian :

    0.0

    prime_meridian_name :

    Greenwich

    geographic_crs_name :

    WGS 84

    horizontal_datum_name :

    World Geodetic System 1984

    grid_mapping_name :

    latitude_longitude

    spatial_ref :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    GeoTransform :

    113.78465630535715 0.00026949458523585647 0.0 22.573674943111044 0.0 -0.00026949458523585647

    array(0)

• Data variables: (1)
  • band_data
    (y, x)
    float32
    -0.0636 -0.06632 ... -0.02509

    array([[-0.06360051, -0.06632194, -0.06632194, ..., -0.01780916,
            -0.00751405, -0.01262642],
           [-0.05900084, -0.08287107, -0.08287107, ..., -0.01780916,
            -0.00667491, -0.01780916],
           [-0.07827458, -0.07091723, -0.06632194, ..., -0.01268844,
            -0.02281629, -0.00155308],
           ...,
           [-0.08128203, -0.06840537, -0.08617212, ..., -0.02509378,
            -0.03017435, -0.02509378],
           [-0.08128203, -0.06329114, -0.07134316, ..., -0.03017435,
            -0.02509378, -0.03017435],
           [-0.07324164, -0.06329114, -0.08128203, ..., -0.03017435,
            -0.035264  , -0.02509378]], dtype=float32)

• Indexes: (2)
  • x
    PandasIndex

    PandasIndex(Float64Index([113.83518654008887, 113.83545603467411, 113.83572552925935,
                  113.83599502384457, 113.83626451842981, 113.83653401301505,
                  113.83680350760028, 113.83707300218552, 113.83734249677076,
                    113.837611991356,
                  ...
                  114.43939340018765,  114.4396628947729, 114.43993238935813,
                  114.44020188394337,  114.4404713785286, 114.44074087311384,
                  114.44101036769908, 114.44127986228432, 114.44154935686954,
                  114.44181885145478],
                 dtype='float64', name='x', length=2252))

  • y
    PandasIndex

    PandasIndex(Float64Index([ 22.56168243406805, 22.561412939482814, 22.561143444897578,
                   22.56087395031234, 22.560604455727105,  22.56033496114187,
                  22.560065466556633,   22.5597959719714, 22.559526477386164,
                  22.559256982800928,
                  ...
                   22.15582358870285, 22.155554094117615,  22.15528459953238,
                  22.155015104947143, 22.154745610361907,  22.15447611577667,
                  22.154206621191435,   22.1539371266062, 22.153667632020962,
                  22.153398137435726],
                 dtype='float64', name='y', length=1516))

• Attributes: (0)

Plot the 2011 mean NDVI with city limits

The Band 15 is the calibrated Land Surface Temperature Band in Kelvin, so we have to subtract 273.15 from the original band data to convert it to Celsius

<xarray.Dataset>
Dimensions:      (x: 5, y: 5)
Coordinates:
    band         int64 15
  * x            (x) float64 113.8 113.8 113.8 113.8 113.8
  * y            (y) float64 22.56 22.56 22.56 22.56 22.56
    spatial_ref  int64 0
Data variables:
    band_data    (y, x) float32 19.86 19.86 19.86 20.4 ... 17.04 17.04 17.04

xarray.Dataset

• Dimensions:
  • x: 5
  • y: 5
• Coordinates: (4)
  • band
    ()
    int64
    15

    array(15)

  • x
    (x)
    float64
    113.8 113.8 113.8 113.8 113.8

    array([113.835187, 113.835456, 113.835726, 113.835995, 113.836265])

  • y
    (y)
    float64
    22.56 22.56 22.56 22.56 22.56

    array([22.561682, 22.561413, 22.561143, 22.560874, 22.560604])

  • spatial_ref
    ()
    int64
    0

    crs_wkt :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    semi_major_axis :

    6378137.0

    semi_minor_axis :

    6356752.314245179

    inverse_flattening :

    298.257223563

    reference_ellipsoid_name :

    WGS 84

    longitude_of_prime_meridian :

    0.0

    prime_meridian_name :

    Greenwich

    geographic_crs_name :

    WGS 84

    horizontal_datum_name :

    World Geodetic System 1984

    grid_mapping_name :

    latitude_longitude

    spatial_ref :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    GeoTransform :

    113.78465630535715 0.00026949458523585647 0.0 22.573674943111044 0.0 -0.00026949458523585647

    array(0)

• Data variables: (1)
  • band_data
    (y, x)
    float32
    19.86 19.86 19.86 ... 17.04 17.04

    array([[19.857178, 19.857178, 19.857178, 20.401917, 19.857178],
           [19.857544, 19.857544, 19.857544, 20.402283, 20.402283],
           [19.85791 , 19.85791 , 20.40268 , 20.40268 , 19.85791 ],
           [17.039246, 17.039246, 17.039246, 17.039246, 17.039246],
           [17.039246, 17.039246, 17.039246, 17.039246, 17.039246]],
          dtype=float32)

• Indexes: (2)
  • x
    PandasIndex

    PandasIndex(Float64Index([113.83518654008887, 113.83545603467411, 113.83572552925935,
                  113.83599502384457, 113.83626451842981],
                 dtype='float64', name='x'))

  • y
    PandasIndex

    PandasIndex(Float64Index([ 22.56168243406805, 22.561412939482814, 22.561143444897578,
                   22.56087395031234, 22.560604455727105],
                 dtype='float64', name='y'))

• Attributes: (0)

<xarray.Dataset>
Dimensions:      ()
Coordinates:
    band         int64 15
    spatial_ref  int64 0
Data variables:
    band_data    float32 20.17

xarray.Dataset

• Dimensions:
• Coordinates: (2)
  • band
    ()
    int64
    15

    array(15)

  • spatial_ref
    ()
    int64
    0

    crs_wkt :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    semi_major_axis :

    6378137.0

    semi_minor_axis :

    6356752.314245179

    inverse_flattening :

    298.257223563

    reference_ellipsoid_name :

    WGS 84

    longitude_of_prime_meridian :

    0.0

    prime_meridian_name :

    Greenwich

    geographic_crs_name :

    WGS 84

    horizontal_datum_name :

    World Geodetic System 1984

    grid_mapping_name :

    latitude_longitude

    spatial_ref :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    GeoTransform :

    113.78465630535715 0.00026949458523585647 0.0 22.573674943111044 0.0 -0.00026949458523585647

    array(0)

• Data variables: (1)
  • band_data
    ()
    float32
    20.17

    array(20.173859, dtype=float32)

• Indexes: (0)
• Attributes: (0)

Using Zonal Statistics to calculate the mean temperature and NDVI for each Hong Kong District

	OBJECTID	ID	CNAME	CNAME_S	ENAME	Shape__Area	Shape__Length	geometry	mean_ndvi_2011	max_temp_2016	mean_ndvi_2016	mean_temp_2011
0	1	1	黃大仙區	黄大仙区	WONG TAI SIN	1.092784e+07	17995.640782	POLYGON ((114.17942 22.34905, 114.17946 22.349...	0.211102	24.699992	0.234139	20.704102
1	2	6	九龍城區	九龙城区	KOWLOON CITY	1.184286e+07	31834.409404	MULTIPOLYGON (((114.17700 22.34904, 114.17702 ...	0.090806	26.748404	0.097568	23.277648
2	3	7	觀塘區	观塘区	KWUN TONG	1.322124e+07	25496.700164	POLYGON ((114.24371 22.28620, 114.24370 22.286...	0.121019	26.426119	0.136822	22.402509
3	4	8	西貢區	西贡区	SAI KUNG	1.602944e+08	365545.476363	MULTIPOLYGON (((114.22112 22.35318, 114.22114 ...	0.262977	24.956479	0.282872	21.255087
4	5	11	北區	北区	NORTH	1.619184e+08	192815.532996	MULTIPOLYGON (((114.33576 22.51003, 114.33576 ...	0.251814	22.677352	0.271825	19.671759
5	6	13	中西區	中西区	CENTRAL & WESTERN	1.460458e+07	26733.700864	MULTIPOLYGON (((114.14562 22.29045, 114.14990 ...	0.194695	23.794897	0.199246	18.948752
6	7	14	灣仔區	湾仔区	WAN CHAI	1.161490e+07	19548.075298	MULTIPOLYGON (((114.20012 22.27387, 114.20047 ...	0.191913	24.292477	0.201838	19.608638
7	8	15	東區	东区	EASTERN	2.184211e+07	36188.456666	POLYGON ((114.24738 22.25339, 114.24733 22.253...	0.213661	23.827175	0.236692	19.821950
8	9	17	屯門區	屯门区	TUEN MUN	9.921091e+07	99074.932441	MULTIPOLYGON (((113.93745 22.42638, 113.93765 ...	0.188711	25.061902	0.221932	20.773037
9	10	18	元朗區	元朗区	YUEN LONG	1.648768e+08	91328.856191	MULTIPOLYGON (((113.93832 22.42696, 113.93843 ...	0.196683	23.764657	0.213754	20.540728

Mean 2011 annual temperature for each district

Mean annual NDVI for each district

2016 version

<xarray.Dataset>
Dimensions:      (band: 16, x: 2503, y: 1783)
Coordinates:
  * band         (band) int64 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
  * x            (x) float64 113.8 113.8 113.8 113.8 ... 114.5 114.5 114.5 114.5
  * y            (y) float64 22.57 22.57 22.57 22.57 ... 22.09 22.09 22.09 22.09
    spatial_ref  int64 ...
Data variables:
    band_data    (band, y, x) float32 ...

xarray.Dataset

• Dimensions:
  • band: 16
  • x: 2503
  • y: 1783
• Coordinates: (4)
  • band
    (band)
    int64
    1 2 3 4 5 6 7 ... 11 12 13 14 15 16

    array([ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16])

  • x
    (x)
    float64
    113.8 113.8 113.8 ... 114.5 114.5

    array([113.784791, 113.785061, 113.78533 , ..., 114.458528, 114.458797,
           114.459067])

  • y
    (y)
    float64
    22.57 22.57 22.57 ... 22.09 22.09

    array([22.57354 , 22.573271, 22.573001, ..., 22.09384 , 22.09357 , 22.093301])

  • spatial_ref
    ()
    int64
    ...

    crs_wkt :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    semi_major_axis :

    6378137.0

    semi_minor_axis :

    6356752.314245179

    inverse_flattening :

    298.257223563

    reference_ellipsoid_name :

    WGS 84

    longitude_of_prime_meridian :

    0.0

    prime_meridian_name :

    Greenwich

    geographic_crs_name :

    WGS 84

    horizontal_datum_name :

    World Geodetic System 1984

    grid_mapping_name :

    latitude_longitude

    spatial_ref :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    GeoTransform :

    113.78465630535715 0.00026949458523585647 0.0 22.573674943111044 0.0 -0.00026949458523585647

    [1 values with dtype=int64]

• Data variables: (1)
  • band_data
    (band, y, x)
    float32
    ...

    AREA_OR_POINT :

    Area

    long_name :

    ('SR_B1', 'SR_B2', 'SR_B3', 'SR_B4', 'SR_B5', 'SR_B7', 'QA_PIXEL', 'NDVI', 'FVC', 'TPW', 'TPWpos', 'EM', 'B6_VCID_1', 'B6_VCID_2', 'LST', 'BBE')

    [71405584 values with dtype=float32]

• Indexes: (3)
  • band
    PandasIndex

    PandasIndex(Int64Index([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16], dtype='int64', name='band'))

  • x
    PandasIndex

    PandasIndex(Float64Index([113.78479105264977,   113.785060547235, 113.78533004182023,
                  113.78559953640547, 113.78586903099071, 113.78613852557595,
                  113.78640802016118, 113.78667751474642, 113.78694700933166,
                   113.7872165039169,
                  ...
                  114.45664105364276,   114.456910548228, 114.45718004281322,
                  114.45744953739846,  114.4577190319837, 114.45798852656894,
                  114.45825802115417,  114.4585275157394, 114.45879701032464,
                  114.45906650490988],
                 dtype='float64', name='x', length=2503))

  • y
    PandasIndex

    PandasIndex(Float64Index([22.573540195818428,  22.57327070123319, 22.573001206647955,
                   22.57273171206272, 22.572462217477483, 22.572192722892247,
                  22.571923228307014, 22.571653733721778, 22.571384239136542,
                  22.571114744551306,
                  ...
                  22.095726296195256,  22.09545680161002, 22.095187307024784,
                  22.094917812439547,  22.09464831785431, 22.094378823269075,
                   22.09410932868384, 22.093839834098603, 22.093570339513366,
                   22.09330084492813],
                 dtype='float64', name='y', length=1783))

• Attributes: (0)

<xarray.Dataset>
Dimensions:      (band: 16, x: 2252, y: 1516)
Coordinates:
  * band         (band) int64 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
  * x            (x) float64 113.8 113.8 113.8 113.8 ... 114.4 114.4 114.4 114.4
  * y            (y) float64 22.56 22.56 22.56 22.56 ... 22.15 22.15 22.15 22.15
    spatial_ref  int64 0
Data variables:
    band_data    (band, y, x) float32 1.061e+04 1.032e+04 ... 0.9894 0.9897

xarray.Dataset

• Dimensions:
  • band: 16
  • x: 2252
  • y: 1516
• Coordinates: (4)
  • band
    (band)
    int64
    1 2 3 4 5 6 7 ... 11 12 13 14 15 16

    array([ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16])

  • x
    (x)
    float64
    113.8 113.8 113.8 ... 114.4 114.4

    array([113.835187, 113.835456, 113.835726, ..., 114.44128 , 114.441549,
           114.441819])

  • y
    (y)
    float64
    22.56 22.56 22.56 ... 22.15 22.15

    array([22.561682, 22.561413, 22.561143, ..., 22.153937, 22.153668, 22.153398])

  • spatial_ref
    ()
    int64
    0

    crs_wkt :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    semi_major_axis :

    6378137.0

    semi_minor_axis :

    6356752.314245179

    inverse_flattening :

    298.257223563

    reference_ellipsoid_name :

    WGS 84

    longitude_of_prime_meridian :

    0.0

    prime_meridian_name :

    Greenwich

    geographic_crs_name :

    WGS 84

    horizontal_datum_name :

    World Geodetic System 1984

    grid_mapping_name :

    latitude_longitude

    spatial_ref :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    GeoTransform :

    113.78465630535715 0.00026949458523585647 0.0 22.573674943111044 0.0 -0.00026949458523585647

    array(0)

• Data variables: (1)
  • band_data
    (band, y, x)
    float32
    1.061e+04 1.032e+04 ... 0.9897

    AREA_OR_POINT :

    Area

    long_name :

    ('SR_B1', 'SR_B2', 'SR_B3', 'SR_B4', 'SR_B5', 'SR_B7', 'QA_PIXEL', 'NDVI', 'FVC', 'TPW', 'TPWpos', 'EM', 'B6_VCID_1', 'B6_VCID_2', 'LST', 'BBE')

    array([[[1.0608000e+04, 1.0323000e+04, 1.0323000e+04, ...,
             9.0600000e+03, 9.1580000e+03, 9.0600000e+03],
            [1.0513000e+04, 1.0513000e+04, 1.0513000e+04, ...,
             9.2560000e+03, 8.9630000e+03, 9.0600000e+03],
            [1.0419000e+04, 1.0419000e+04, 1.0512000e+04, ...,
             9.0600000e+03, 9.2560000e+03, 9.1580000e+03],
            ...,
            [1.1319000e+04, 1.1126000e+04, 1.0931000e+04, ...,
             9.3770000e+03, 9.1780000e+03, 9.5740000e+03],
            [1.1028000e+04, 1.1126000e+04, 1.0931000e+04, ...,
             9.3770000e+03, 9.1780000e+03, 9.0780000e+03],
            [1.1223000e+04, 1.1028000e+04, 1.0931000e+04, ...,
             9.3770000e+03, 9.2770000e+03, 9.2770000e+03]],
    
           [[1.1398000e+04, 1.1499000e+04, 1.1499000e+04, ...,
             8.8280000e+03, 8.8280000e+03, 8.8290000e+03],
            [1.1499000e+04, 1.1499000e+04, 1.1499000e+04, ...,
             8.8280000e+03, 8.7240000e+03, 8.7240000e+03],
            [1.1399000e+04, 1.1499000e+04, 1.1702000e+04, ...,
             8.9320000e+03, 8.9320000e+03, 8.8280000e+03],
    ...
            [2.9249011e+02, 2.9249011e+02, 2.9249011e+02, ...,
             2.9369949e+02, 2.9312128e+02, 2.9369949e+02],
            [2.9181238e+02, 2.9181238e+02, 2.9249011e+02, ...,
             2.9369949e+02, 2.9312128e+02, 2.9369949e+02],
            [2.9249011e+02, 2.9249011e+02, 2.9249011e+02, ...,
             2.9369949e+02, 2.9369949e+02, 2.9369949e+02]],
    
           [[9.8959720e-01, 9.8956382e-01, 9.8956382e-01, ...,
             9.8927397e-01, 9.8907739e-01, 9.8888111e-01],
            [9.8964369e-01, 9.8959720e-01, 9.8959720e-01, ...,
             9.8933840e-01, 9.8933840e-01, 9.8891693e-01],
            [9.8978209e-01, 9.8964369e-01, 9.8956239e-01, ...,
             9.8907524e-01, 9.8907524e-01, 9.8888111e-01],
            ...,
            [9.8929220e-01, 9.8937589e-01, 9.8937589e-01, ...,
             9.8950535e-01, 9.8867345e-01, 9.8939520e-01],
            [9.8927212e-01, 9.8929220e-01, 9.8940676e-01, ...,
             9.8939717e-01, 9.8921460e-01, 9.8973274e-01],
            [9.8910576e-01, 9.8927116e-01, 9.8937589e-01, ...,
             9.8888195e-01, 9.8939717e-01, 9.8973274e-01]]], dtype=float32)

• Indexes: (3)
  • band
    PandasIndex

    PandasIndex(Int64Index([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16], dtype='int64', name='band'))

  • x
    PandasIndex

    PandasIndex(Float64Index([113.83518654008887, 113.83545603467411, 113.83572552925935,
                  113.83599502384457, 113.83626451842981, 113.83653401301505,
                  113.83680350760028, 113.83707300218552, 113.83734249677076,
                    113.837611991356,
                  ...
                  114.43939340018765,  114.4396628947729, 114.43993238935813,
                  114.44020188394337,  114.4404713785286, 114.44074087311384,
                  114.44101036769908, 114.44127986228432, 114.44154935686954,
                  114.44181885145478],
                 dtype='float64', name='x', length=2252))

  • y
    PandasIndex

    PandasIndex(Float64Index([ 22.56168243406805, 22.561412939482814, 22.561143444897578,
                   22.56087395031234, 22.560604455727105,  22.56033496114187,
                  22.560065466556633,   22.5597959719714, 22.559526477386164,
                  22.559256982800928,
                  ...
                   22.15582358870285, 22.155554094117615,  22.15528459953238,
                  22.155015104947143, 22.154745610361907,  22.15447611577667,
                  22.154206621191435,   22.1539371266062, 22.153667632020962,
                  22.153398137435726],
                 dtype='float64', name='y', length=1516))

• Attributes: (0)

<xarray.Dataset>
Dimensions:      ()
Coordinates:
    spatial_ref  int64 0
Data variables:
    band_data    float32 0.09237

xarray.Dataset

• Dimensions:
• Coordinates: (1)
  • spatial_ref
    ()
    int64
    0

    crs_wkt :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    semi_major_axis :

    6378137.0

    semi_minor_axis :

    6356752.314245179

    inverse_flattening :

    298.257223563

    reference_ellipsoid_name :

    WGS 84

    longitude_of_prime_meridian :

    0.0

    prime_meridian_name :

    Greenwich

    geographic_crs_name :

    WGS 84

    horizontal_datum_name :

    World Geodetic System 1984

    grid_mapping_name :

    latitude_longitude

    spatial_ref :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    GeoTransform :

    113.78465630535715 0.00026949458523585647 0.0 22.573674943111044 0.0 -0.00026949458523585647

    array(0)

• Data variables: (1)
  • band_data
    ()
    float32
    0.09237

    array(0.09237264, dtype=float32)

• Indexes: (0)
• Attributes: (0)

<xarray.Dataset>
Dimensions:      ()
Coordinates:
    band         int64 15
    spatial_ref  int64 0
Data variables:
    band_data    float32 23.15

xarray.Dataset

• Dimensions:
• Coordinates: (2)
  • band
    ()
    int64
    15

    array(15)

  • spatial_ref
    ()
    int64
    0

    crs_wkt :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    semi_major_axis :

    6378137.0

    semi_minor_axis :

    6356752.314245179

    inverse_flattening :

    298.257223563

    reference_ellipsoid_name :

    WGS 84

    longitude_of_prime_meridian :

    0.0

    prime_meridian_name :

    Greenwich

    geographic_crs_name :

    WGS 84

    horizontal_datum_name :

    World Geodetic System 1984

    grid_mapping_name :

    latitude_longitude

    spatial_ref :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    GeoTransform :

    113.78465630535715 0.00026949458523585647 0.0 22.573674943111044 0.0 -0.00026949458523585647

    array(0)

• Data variables: (1)
  • band_data
    ()
    float32
    23.15

    array(23.14934, dtype=float32)

• Indexes: (0)
• Attributes: (0)

	OBJECTID	ID	CNAME	CNAME_S	ENAME	Shape__Area	Shape__Length	geometry	mean_ndvi_2011	mean_ndvi_2016	mean_temp_2011	mean_temp_2016
0	1	1	黃大仙區	黄大仙区	WONG TAI SIN	1.092784e+07	17995.640782	POLYGON ((114.17942 22.34905, 114.17946 22.349...	0.211102	0.234139	20.704102	24.699992
1	2	6	九龍城區	九龙城区	KOWLOON CITY	1.184286e+07	31834.409404	MULTIPOLYGON (((114.17700 22.34904, 114.17702 ...	0.090806	0.097568	23.277648	26.748404
2	3	7	觀塘區	观塘区	KWUN TONG	1.322124e+07	25496.700164	POLYGON ((114.24371 22.28620, 114.24370 22.286...	0.121019	0.136822	22.402509	26.426119
3	4	8	西貢區	西贡区	SAI KUNG	1.602944e+08	365545.476363	MULTIPOLYGON (((114.22112 22.35318, 114.22114 ...	0.262977	0.282872	21.255087	24.956479
4	5	11	北區	北区	NORTH	1.619184e+08	192815.532996	MULTIPOLYGON (((114.33576 22.51003, 114.33576 ...	0.251814	0.271825	19.671759	22.677352
5	6	13	中西區	中西区	CENTRAL & WESTERN	1.460458e+07	26733.700864	MULTIPOLYGON (((114.14562 22.29045, 114.14990 ...	0.194695	0.199246	18.948752	23.794897
6	7	14	灣仔區	湾仔区	WAN CHAI	1.161490e+07	19548.075298	MULTIPOLYGON (((114.20012 22.27387, 114.20047 ...	0.191913	0.201838	19.608638	24.292477
7	8	15	東區	东区	EASTERN	2.184211e+07	36188.456666	POLYGON ((114.24738 22.25339, 114.24733 22.253...	0.213661	0.236692	19.821950	23.827175
8	9	17	屯門區	屯门区	TUEN MUN	9.921091e+07	99074.932441	MULTIPOLYGON (((113.93745 22.42638, 113.93765 ...	0.188711	0.221932	20.773037	25.061902
9	10	18	元朗區	元朗区	YUEN LONG	1.648768e+08	91328.856191	MULTIPOLYGON (((113.93832 22.42696, 113.93843 ...	0.196683	0.213754	20.540728	23.764657

2021 Version

<xarray.Dataset>
Dimensions:      (band: 16, x: 2503, y: 1783)
Coordinates:
  * band         (band) int64 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
  * x            (x) float64 113.8 113.8 113.8 113.8 ... 114.5 114.5 114.5 114.5
  * y            (y) float64 22.57 22.57 22.57 22.57 ... 22.09 22.09 22.09 22.09
    spatial_ref  int64 ...
Data variables:
    band_data    (band, y, x) float32 ...

xarray.Dataset

• Dimensions:
  • band: 16
  • x: 2503
  • y: 1783
• Coordinates: (4)
  • band
    (band)
    int64
    1 2 3 4 5 6 7 ... 11 12 13 14 15 16

    array([ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16])

  • x
    (x)
    float64
    113.8 113.8 113.8 ... 114.5 114.5

    array([113.784791, 113.785061, 113.78533 , ..., 114.458528, 114.458797,
           114.459067])

  • y
    (y)
    float64
    22.57 22.57 22.57 ... 22.09 22.09

    array([22.57354 , 22.573271, 22.573001, ..., 22.09384 , 22.09357 , 22.093301])

  • spatial_ref
    ()
    int64
    ...

    crs_wkt :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    semi_major_axis :

    6378137.0

    semi_minor_axis :

    6356752.314245179

    inverse_flattening :

    298.257223563

    reference_ellipsoid_name :

    WGS 84

    longitude_of_prime_meridian :

    0.0

    prime_meridian_name :

    Greenwich

    geographic_crs_name :

    WGS 84

    horizontal_datum_name :

    World Geodetic System 1984

    grid_mapping_name :

    latitude_longitude

    spatial_ref :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    GeoTransform :

    113.78465630535715 0.00026949458523585647 0.0 22.573674943111044 0.0 -0.00026949458523585647

    [1 values with dtype=int64]

• Data variables: (1)
  • band_data
    (band, y, x)
    float32
    ...

    AREA_OR_POINT :

    Area

    long_name :

    ('SR_B1', 'SR_B2', 'SR_B3', 'SR_B4', 'SR_B5', 'SR_B7', 'QA_PIXEL', 'NDVI', 'FVC', 'TPW', 'TPWpos', 'EM', 'B6_VCID_1', 'B6_VCID_2', 'LST', 'BBE')

    [71405584 values with dtype=float32]

• Indexes: (3)
  • band
    PandasIndex

    PandasIndex(Int64Index([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16], dtype='int64', name='band'))

  • x
    PandasIndex

    PandasIndex(Float64Index([113.78479105264977,   113.785060547235, 113.78533004182023,
                  113.78559953640547, 113.78586903099071, 113.78613852557595,
                  113.78640802016118, 113.78667751474642, 113.78694700933166,
                   113.7872165039169,
                  ...
                  114.45664105364276,   114.456910548228, 114.45718004281322,
                  114.45744953739846,  114.4577190319837, 114.45798852656894,
                  114.45825802115417,  114.4585275157394, 114.45879701032464,
                  114.45906650490988],
                 dtype='float64', name='x', length=2503))

  • y
    PandasIndex

    PandasIndex(Float64Index([22.573540195818428,  22.57327070123319, 22.573001206647955,
                   22.57273171206272, 22.572462217477483, 22.572192722892247,
                  22.571923228307014, 22.571653733721778, 22.571384239136542,
                  22.571114744551306,
                  ...
                  22.095726296195256,  22.09545680161002, 22.095187307024784,
                  22.094917812439547,  22.09464831785431, 22.094378823269075,
                   22.09410932868384, 22.093839834098603, 22.093570339513366,
                   22.09330084492813],
                 dtype='float64', name='y', length=1783))

• Attributes: (0)

<xarray.Dataset>
Dimensions:      (band: 16, x: 2252, y: 1516)
Coordinates:
  * band         (band) int64 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
  * x            (x) float64 113.8 113.8 113.8 113.8 ... 114.4 114.4 114.4 114.4
  * y            (y) float64 22.56 22.56 22.56 22.56 ... 22.15 22.15 22.15 22.15
    spatial_ref  int64 0
Data variables:
    band_data    (band, y, x) float32 9.098e+03 9.027e+03 ... 0.9903 0.9903

xarray.Dataset

• Dimensions:
  • band: 16
  • x: 2252
  • y: 1516
• Coordinates: (4)
  • band
    (band)
    int64
    1 2 3 4 5 6 7 ... 11 12 13 14 15 16

    array([ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16])

  • x
    (x)
    float64
    113.8 113.8 113.8 ... 114.4 114.4

    array([113.835187, 113.835456, 113.835726, ..., 114.44128 , 114.441549,
           114.441819])

  • y
    (y)
    float64
    22.56 22.56 22.56 ... 22.15 22.15

    array([22.561682, 22.561413, 22.561143, ..., 22.153937, 22.153668, 22.153398])

  • spatial_ref
    ()
    int64
    0

    crs_wkt :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    semi_major_axis :

    6378137.0

    semi_minor_axis :

    6356752.314245179

    inverse_flattening :

    298.257223563

    reference_ellipsoid_name :

    WGS 84

    longitude_of_prime_meridian :

    0.0

    prime_meridian_name :

    Greenwich

    geographic_crs_name :

    WGS 84

    horizontal_datum_name :

    World Geodetic System 1984

    grid_mapping_name :

    latitude_longitude

    spatial_ref :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    GeoTransform :

    113.78465630535715 0.00026949458523585647 0.0 22.573674943111044 0.0 -0.00026949458523585647

    array(0)

• Data variables: (1)
  • band_data
    (band, y, x)
    float32
    9.098e+03 9.027e+03 ... 0.9903

    AREA_OR_POINT :

    Area

    long_name :

    ('SR_B1', 'SR_B2', 'SR_B3', 'SR_B4', 'SR_B5', 'SR_B7', 'QA_PIXEL', 'NDVI', 'FVC', 'TPW', 'TPWpos', 'EM', 'B6_VCID_1', 'B6_VCID_2', 'LST', 'BBE')

    array([[[9.0980000e+03, 9.0270000e+03, 9.0270000e+03, ...,
             9.4170000e+03, 9.4170000e+03, 9.2980000e+03],
            [9.1690000e+03, 9.1690000e+03, 9.1690000e+03, ...,
             9.1790000e+03, 9.2980000e+03, 9.4140000e+03],
            [8.9550000e+03, 9.0980000e+03, 9.0980000e+03, ...,
             9.2970000e+03, 9.2970000e+03, 9.5330000e+03],
            ...,
            [9.7300000e+03, 9.7300000e+03, 9.6170000e+03, ...,
             9.0890000e+03, 9.3290000e+03, 9.5690000e+03],
            [9.7300000e+03, 9.8430000e+03, 9.7300000e+03, ...,
             9.5680000e+03, 9.4490000e+03, 9.3290000e+03],
            [1.0111000e+04, 1.0068000e+04, 9.6180000e+03, ...,
             9.6880000e+03, 9.4490000e+03, 9.3290000e+03]],
    
           [[9.7320000e+03, 9.8100000e+03, 9.8100000e+03, ...,
             9.1870000e+03, 9.1870000e+03, 9.3100000e+03],
            [9.8100000e+03, 9.7320000e+03, 9.7320000e+03, ...,
             9.4340000e+03, 9.3100000e+03, 9.3090000e+03],
            [9.8890000e+03, 9.8100000e+03, 9.6530000e+03, ...,
             9.3090000e+03, 9.1850000e+03, 9.4320000e+03],
    ...
            [2.8975958e+02, 2.8975958e+02, 2.8975958e+02, ...,
             2.9418109e+02, 2.9418109e+02, 2.9418109e+02],
            [2.8975958e+02, 2.8975958e+02, 2.8975958e+02, ...,
             2.9418109e+02, 2.9363739e+02, 2.9418109e+02],
            [2.9863339e+02, 2.8906863e+02, 2.8906863e+02, ...,
             2.9418109e+02, 2.9418109e+02, 2.9418109e+02]],
    
           [[9.8935848e-01, 9.9028999e-01, 9.9028999e-01, ...,
             9.9028999e-01, 9.8886591e-01, 9.9028999e-01],
            [9.8974252e-01, 9.9028999e-01, 9.9028999e-01, ...,
             9.8911208e-01, 9.8936027e-01, 9.8984039e-01],
            [9.9028999e-01, 9.9028999e-01, 9.8948997e-01, ...,
             9.9028999e-01, 9.8936188e-01, 9.8984039e-01],
            ...,
            [9.8903477e-01, 9.8892921e-01, 9.8904902e-01, ...,
             9.9028999e-01, 9.9028999e-01, 9.9028999e-01],
            [9.8903584e-01, 9.8882341e-01, 9.8882341e-01, ...,
             9.9028999e-01, 9.9028999e-01, 9.9028999e-01],
            [9.9028999e-01, 9.8882437e-01, 9.8903435e-01, ...,
             9.9028999e-01, 9.9028999e-01, 9.9028999e-01]]], dtype=float32)

• Indexes: (3)
  • band
    PandasIndex

    PandasIndex(Int64Index([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16], dtype='int64', name='band'))

  • x
    PandasIndex

    PandasIndex(Float64Index([113.83518654008887, 113.83545603467411, 113.83572552925935,
                  113.83599502384457, 113.83626451842981, 113.83653401301505,
                  113.83680350760028, 113.83707300218552, 113.83734249677076,
                    113.837611991356,
                  ...
                  114.43939340018765,  114.4396628947729, 114.43993238935813,
                  114.44020188394337,  114.4404713785286, 114.44074087311384,
                  114.44101036769908, 114.44127986228432, 114.44154935686954,
                  114.44181885145478],
                 dtype='float64', name='x', length=2252))

  • y
    PandasIndex

    PandasIndex(Float64Index([ 22.56168243406805, 22.561412939482814, 22.561143444897578,
                   22.56087395031234, 22.560604455727105,  22.56033496114187,
                  22.560065466556633,   22.5597959719714, 22.559526477386164,
                  22.559256982800928,
                  ...
                   22.15582358870285, 22.155554094117615,  22.15528459953238,
                  22.155015104947143, 22.154745610361907,  22.15447611577667,
                  22.154206621191435,   22.1539371266062, 22.153667632020962,
                  22.153398137435726],
                 dtype='float64', name='y', length=1516))

• Attributes: (0)

<xarray.Dataset>
Dimensions:      ()
Coordinates:
    spatial_ref  int64 0
Data variables:
    band_data    float32 0.5408

xarray.Dataset

• Dimensions:
• Coordinates: (1)
  • spatial_ref
    ()
    int64
    0

    crs_wkt :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    semi_major_axis :

    6378137.0

    semi_minor_axis :

    6356752.314245179

    inverse_flattening :

    298.257223563

    reference_ellipsoid_name :

    WGS 84

    longitude_of_prime_meridian :

    0.0

    prime_meridian_name :

    Greenwich

    geographic_crs_name :

    WGS 84

    horizontal_datum_name :

    World Geodetic System 1984

    grid_mapping_name :

    latitude_longitude

    spatial_ref :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    GeoTransform :

    113.78465630535715 0.00026949458523585647 0.0 22.573674943111044 0.0 -0.00026949458523585647

    array(0)

• Data variables: (1)
  • band_data
    ()
    float32
    0.5408

    array(0.54077834, dtype=float32)

• Indexes: (0)
• Attributes: (0)

<xarray.Dataset>
Dimensions:      ()
Coordinates:
    band         int64 15
    spatial_ref  int64 0
Data variables:
    band_data    float32 20.91

xarray.Dataset

• Dimensions:
• Coordinates: (2)
  • band
    ()
    int64
    15

    array(15)

  • spatial_ref
    ()
    int64
    0

    crs_wkt :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    semi_major_axis :

    6378137.0

    semi_minor_axis :

    6356752.314245179

    inverse_flattening :

    298.257223563

    reference_ellipsoid_name :

    WGS 84

    longitude_of_prime_meridian :

    0.0

    prime_meridian_name :

    Greenwich

    geographic_crs_name :

    WGS 84

    horizontal_datum_name :

    World Geodetic System 1984

    grid_mapping_name :

    latitude_longitude

    spatial_ref :

    GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]]

    GeoTransform :

    113.78465630535715 0.00026949458523585647 0.0 22.573674943111044 0.0 -0.00026949458523585647

    array(0)

• Data variables: (1)
  • band_data
    ()
    float32
    20.91

    array(20.90542, dtype=float32)

• Indexes: (0)
• Attributes: (0)

	OBJECTID	ID	CNAME	CNAME_S	ENAME	Shape__Area	Shape__Length	geometry	mean_ndvi_2011	mean_ndvi_2016	mean_temp_2011	mean_temp_2016	mean_temp_2021	mean_ndvi_2021
0	1	1	黃大仙區	黄大仙区	WONG TAI SIN	1.092784e+07	17995.640782	POLYGON ((114.17942 22.34905, 114.17946 22.349...	0.211102	0.234139	20.704102	24.699992	21.227092	0.215096
1	2	6	九龍城區	九龙城区	KOWLOON CITY	1.184286e+07	31834.409404	MULTIPOLYGON (((114.17700 22.34904, 114.17702 ...	0.090806	0.097568	23.277648	26.748404	22.868930	0.081260
2	3	7	觀塘區	观塘区	KWUN TONG	1.322124e+07	25496.700164	POLYGON ((114.24371 22.28620, 114.24370 22.286...	0.121019	0.136822	22.402509	26.426119	22.577353	0.117286
3	4	8	西貢區	西贡区	SAI KUNG	1.602944e+08	365545.476363	MULTIPOLYGON (((114.22112 22.35318, 114.22114 ...	0.262977	0.282872	21.255087	24.956479	21.121994	0.278582
4	5	11	北區	北区	NORTH	1.619184e+08	192815.532996	MULTIPOLYGON (((114.33576 22.51003, 114.33576 ...	0.251814	0.271825	19.671759	22.677352	20.621217	0.256020
5	6	13	中西區	中西区	CENTRAL & WESTERN	1.460458e+07	26733.700864	MULTIPOLYGON (((114.14562 22.29045, 114.14990 ...	0.194695	0.199246	18.948752	23.794897	20.243429	0.200184
6	7	14	灣仔區	湾仔区	WAN CHAI	1.161490e+07	19548.075298	MULTIPOLYGON (((114.20012 22.27387, 114.20047 ...	0.191913	0.201838	19.608638	24.292477	21.345188	0.192107
7	8	15	東區	东区	EASTERN	2.184211e+07	36188.456666	POLYGON ((114.24738 22.25339, 114.24733 22.253...	0.213661	0.236692	19.821950	23.827175	21.877428	0.212786
8	9	17	屯門區	屯门区	TUEN MUN	9.921091e+07	99074.932441	MULTIPOLYGON (((113.93745 22.42638, 113.93765 ...	0.188711	0.221932	20.773037	25.061902	20.751241	0.195801
9	10	18	元朗區	元朗区	YUEN LONG	1.648768e+08	91328.856191	MULTIPOLYGON (((113.93832 22.42696, 113.93843 ...	0.196683	0.213754	20.540728	23.764657	20.284002	0.203589

Cluster Analysis

We will perform K-means Cluster Analysis for our final analytical output.

Index(['NAME_EN', 'NAME_TC', 'total_pop_2016', 'total_pop_2021',
       'total_pop_2011', 'median_age_2011', 'median_age_2016',
       'median_age_2021', 'OBJECTID', 'ID', 'CNAME', 'CNAME_S', 'ENAME',
       'Shape__Area', 'Shape__Length', 'road_length', 'area', 'road_density',
       'open_space_area', 'mean_temp_2011', 'mean_temp_2016', 'mean_temp_2021',
       'mean_ndvi_2011', 'mean_ndvi_2016', 'mean_ndvi_2021',
       'building_density', 'geometry'],
      dtype='object')

We first import all libraries we need, and then choose the key variables to determine the clustering pattern in our merged dataset, which are: Total population in 2011, 2016, 2021, median age of the population (in three years) for each district in Hong Kong, road density, open space area, mean temperature in three years, mean ndvi (normalized difference vegetation index) in three years, and building density by district.

Initially, we tested a cluster range for our preliminary testing of the clustering model. Random_state set to 42 so that we can make the randomness deterministic.

This ‘Kneed’ package can help us determine the ‘knee’ point quantitatively. It gives out the number of cluster of 4, which aligns with our initial testing and decision for Scree plot.

Scree plot is one way to determine how many clusters we should input for our kmeans model. From the shape of the curve, we probably will select 3 or 4 for our final number of cluster.

4 [2, 3, 4, 5, 6, 7, 8, 9]

After performing the kmeans analysis, we gives each district a ‘label’ so that we can visualize the cluster on the map.

	label	size
0	0	1
1	1	5
2	2	7
3	3	5

We finally plot our result as a map. It can be seen that four clusters are shown quite evenly across Hong Kong, except Cluster 0 (in Dark Red) which demonstrates the Hong Kong Island District with airports, few buildings, low population density, mainly mountains and high volume of vegetations).

It demonstrates that the pale blue regions (Cluster 2) are remote suburb areas with relatively low population and building density, and are having large areas of parks and natural landscapes. The orange regions (Cluster 1) represents the main populated living area where population densities are high and buildings are dense. Dark Blue (Cluster 3) regions are the main business & entertainment and travel areas where there are many buildings and population, while the area is small so density is high.

If we are to assign names to those clusters, the dark red (cluster 0) will be Emergent Mixed-Use Area, the orange (cluster 1) will be Urban Residential Zone, the pale blue (cluster 2) will be Suburb Oasis, the dark blue (cluster 3) will be Commercial and Entertainment Hub, which is also identifed as with high vulnerbility to future urban heat index.

Appendix

(https://code.earthengine.google.com/531a9a4d8b3a6c1a3fafab218f7d7159).

Navigating Data Source Challenges in Urban Heat Index Analysis Acquiring and utilizing data for our Urban Heat Index analysis involved a series of steps and challenges, particularly in data calibration and exportation. Our primary data source was a feature collection from the Google Earth Engine (GEE), which provided us with Landsat 7 satellite imagery. We initially started with the Raw Scene version of the Landsat 7 data. This version had no calibration applied to the digital numbers, meaning the data was in its most unprocessed form.

Subsequently, we discovered the “Top of Atmosphere” (TOA) version of the Landsat 7 data, which had been converted by the US Geological Survey (USGS). This TOA version seemed promising and functioned well within the Google Earth Engine environment. However, we encountered a significant issue when exporting the data from GEE to Python. Specifically, Band 3 and Band 4 values were altered during this export process. This alteration led to the NDVI values falling outside the expected range of -1 to 1, a puzzling issue that we struggled to resolve.

To address this, we implemented cloud masks to remove the impact of cloud cover from our analysis, hoping this would correct the discrepancies in the NDVI values. Unfortunately, this step did not yield the desired results. The problem might lie in exporting the data, particularly the stage where data is transferred to Google Drive. The data, when downloaded, undergoes a zipping and unzipping process on our laptops. We hypothesized that this process could potentially alter the band values for Band 3, 4, and 6 (the Thermal Band similar to Landsat 8, Band 10, and Band 11), resulting in the discrepancies we observed.