Environment & Habitats

What are the general environmental conditions at the mosquito observations?

import pandas as pd
pd.set_option("display.max_columns", None)
import geopandas as gpd
import matplotlib.pyplot as plt
import seaborn as sns
import ee
import geemap
ee.Authenticate()

# Write your project ID here, in quotes
ee.Initialize(project = "emerge-lessons")

mosquito = gpd.read_file('https://github.com/geo-di-lab/emerge-lessons/raw/refs/heads/main/docs/data/globe_mosquito.zip')

For all the Google Earth Engine datasets we looked at in the previous lesson, we’ll add the data as extra columns onto the mosquito dataset.

def mask_s2_clouds(image):
  """Masks clouds in a Sentinel-2 image using the QA band.
  Args:
      image (ee.Image): A Sentinel-2 image.
  Returns:
      ee.Image: A cloud-masked Sentinel-2 image.
  """
  qa = image.select('QA60')

  # Bits 10 and 11 are clouds and cirrus, respectively.
  cloud_bit_mask = 1 << 10
  cirrus_bit_mask = 1 << 11

  # Both flags should be set to zero, indicating clear conditions.
  mask = (
      qa.bitwiseAnd(cloud_bit_mask)
      .eq(0)
      .And(qa.bitwiseAnd(cirrus_bit_mask).eq(0))
  )

  return image.updateMask(mask).divide(10000)

def get_data_for_point(point):
    # Extract the point's date property
    point_date = ee.Date(point.get('MeasuredDate'))
    day_before = point_date.advance(-1, 'day')
    month_before = point_date.advance(-30, 'day')

    #### Land Surface Temperature ####
    lst = (
        ee.ImageCollection('MODIS/061/MOD11A1')
        .filterDate(day_before, point_date)
        .select(['LST_Day_1km', 'QC_Day', 'LST_Night_1km', 'QC_Night', 'Clear_day_cov', 'Clear_night_cov'])
        .median()
        # Sample LST at the point’s location
        .reduceRegion(
            reducer=ee.Reducer.median(),
            geometry=point.geometry(),
            scale=1000
        )
    )
    point = point.set(lst)

    #### Sentinel-2, Median of Previous Month (used to calculate NDVI and NDWI) ####
    sentinel2 = (
        ee.ImageCollection('COPERNICUS/S2_SR_HARMONIZED')
        .filterDate(month_before, point_date)
        .map(mask_s2_clouds)
        .median()
        # Sample median band values at the point’s location
        .reduceRegion(
            reducer=ee.Reducer.median(),
            geometry=point.geometry(),
            scale=10
        )
    )
    point = point.set(sentinel2)

    #### Precipitation ####
    rain = (
        ee.ImageCollection('UCSB-CHG/CHIRPS/DAILY')
        .filterDate(day_before, point_date)
        .select('precipitation')
        .sum()
        # Sample median band values at the point’s location
        .reduceRegion(
            reducer=ee.Reducer.median(),
            geometry=point.geometry(),
            scale=5566
        )
    )
    point = point.set(rain)

    return point

Because the entire dataset takes up a lot of memory, we will use a subset of the data, from 2024 (which has 2039 observations, compared to the 43009 total observations across 2018-2024).

mosquito_2024 = mosquito[mosquito['MeasuredDate'].str.startswith('2024')]

# Apply the function to all points in 2024
points_fc = geemap.geopandas_to_ee(mosquito_2024)
result_fc = points_fc.map(get_data_for_point)

data = geemap.ee_to_gdf(result_fc)

## Calculate NDVI
data['NDVI'] = (data['B8'] - data['B4']) / (data['B8'] + data['B4'])

## Calculate NDWI
data['NDWI'] = (data['B3'] - data['B8']) / (data['B3'] + data['B8'])

data

	geometry	AOT	B1	B11	B12	B2	B3	B4	B5	B6	B7	B8	B8A	B9	BreedingGroundEliminated	CountryCode	CountryName	DataSource	Elevation	ExtraData	GlobeTeams	LarvaeCount	LarvaeCountProcessed	LocationAccuracyM	LocationMethod	MSK_CLASSI_CIRRUS	MSK_CLASSI_OPAQUE	MSK_CLASSI_SNOW_ICE	MSK_CLDPRB	MSK_SNWPRB	MeasuredAt	MeasuredDate	MeasurementElevation	MeasurementLatitude	MeasurementLongitude	MosquitoHabitatMapperId	MosquitoPupae	OrganizationId	OrganizationName	Protocol	QA60	QC_Day	QC_Night	SCL	SiteId	SiteName	TCI_B	TCI_G	TCI_R	Userid	WVP	WaterSource	WaterSourcePhotoUrls	WaterSourceType	precipitation	LastIdentifyStage	Clear_day_cov	LST_Day_1km	Clear_night_cov	LST_Night_1km	MosquitoAdults	MosquitoEggs	Comments	LarvaFullBodyPhotoUrls	Genus	Species	AbdomenCloseupPhotoUrls	QA10	QA20	NDVI	NDWI
0	POINT (-44.26597 -2.56197)	0.0228	0.10210	0.45830	0.39510	0.1128	0.1506	0.2208	0.27940	0.31590	0.3443	0.31560	0.33190	0.30720	false	BRA	Brazil	GLOBE Observer App	6.3	LarvaeVisibleNo	[COLUNSLZ]	0	0.0	13	automatic	0.0	0.0	0.0	0.00000	0.0	2024-12-31T17:16:00	2024-12-31	0	-2.561700	-44.265700	46287	false	17459532	Brazil Citizen Science	mosquito_habitat_mapper	0.0	2.0	2.0	0.00050	371514	23MNT816168	0.01150	0.01540	0.02250	137422629	0.37670	ovitrap	https://data.globe.gov/system/photos/2024/12/3...	container: artificial	0.000000	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	0.176734	-0.353925
1	POINT (-44.26597 -2.56197)	0.0228	0.10210	0.45830	0.39510	0.1128	0.1506	0.2208	0.27940	0.31590	0.3443	0.31560	0.33190	0.30720	false	BRA	Brazil	GLOBE Observer App	6.3	LarvaeVisibleNo	[COLUNSLZ]	0	0.0	13	automatic	0.0	0.0	0.0	0.00000	0.0	2024-12-31T17:20:00	2024-12-31	0	-2.561700	-44.265700	46290	false	17459532	Brazil Citizen Science	mosquito_habitat_mapper	0.0	2.0	2.0	0.00050	371514	23MNT816168	0.01150	0.01540	0.02250	137422629	0.37670	ovitrap	https://data.globe.gov/system/photos/2024/12/3...	container: artificial	0.000000	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	0.176734	-0.353925
2	POINT (-44.30288 -2.51676)	0.0168	0.12640	0.30260	0.25450	0.1234	0.1392	0.1886	0.22630	0.22790	0.2369	0.19860	0.23540	0.23600	true	BRA	Brazil	GLOBE Observer App	7.4	LarvaeVisibleNo	[COLUNSLZ]	0	0.0	51	automatic	0.0	0.0	0.0	0.00000	0.0	2024-12-31T22:32:00	2024-12-31	0	-2.516300	-44.302300	46482	false	17459532	Brazil Citizen Science	mosquito_habitat_mapper	0.0	3.0	3.0	0.00080	372864	23MNT775218	0.01260	0.01420	0.01920	137420190	0.41330	cement, metal or plastic tank	NaN	container: artificial	0.000000	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	0.025826	-0.175844
3	POINT (-44.05526 -2.86396)	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	true	BRA	Brazil	GLOBE Observer App	20.6	LarvaeVisibleNo	[COLUNSLZ]	0	0.0	66	automatic	NaN	NaN	NaN	NaN	NaN	2024-12-31T00:05:00	2024-12-31	0	-2.863900	-44.054900	46203	false	17459532	Brazil Citizen Science	mosquito_habitat_mapper	NaN	2.0	2.0	NaN	373085	23MPS050834	NaN	NaN	NaN	137419937	NaN	can or bottle	NaN	container: artificial	11.092295	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
4	POINT (-44.05526 -2.86396)	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	true	BRA	Brazil	GLOBE Observer App	20.6	LarvaeVisibleNo	[COLUNSLZ]	0	0.0	28	automatic	NaN	NaN	NaN	NaN	NaN	2024-12-31T00:23:00	2024-12-31	0	-2.863900	-44.055000	46223	false	17459532	Brazil Citizen Science	mosquito_habitat_mapper	NaN	2.0	2.0	NaN	373085	23MPS050834	NaN	NaN	NaN	137419937	NaN	lake	NaN	still: lake/pond/swamp	11.092295	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2037	POINT (2.35801 6.37588)	0.0372	0.12305	0.24485	0.21515	0.1320	0.1425	0.1573	0.17280	0.18885	0.2138	0.20650	0.22225	0.20735	true	BEN	Benin	GLOBE Observer App	6.0	LarvaeVisibleYes	NaN	NaN	NaN	1419	automatic	NaN	NaN	NaN	0.00075	0.0	2024-01-04T15:20:00	2024-01-04	0	6.375900	2.358300	44122	false	23502301	University of Abomey Calavi	mosquito_habitat_mapper	0.0	2.0	2.0	0.00065	336012	31NDH290048	0.01345	0.01455	0.01605	51262501	0.29490	puddle, vehicle or animal tracks	https://data.globe.gov/system/photos/2024/01/0...	still: lake/pond/swamp	0.000000	identify-later	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	0.0	0.0	0.135239	-0.183381
2038	POINT (2.37609 6.37862)	0.0375	0.14665	0.25770	0.22050	0.1411	0.1573	0.1552	0.18585	0.21075	0.2359	0.25820	0.24880	0.29900	false	BEN	Benin	GLOBE Observer App	3.3	LarvaeVisibleYes	NaN	NaN	NaN	64	automatic	NaN	NaN	NaN	0.00050	0.0	2024-01-04T15:52:00	2024-01-04	0	6.379400	2.376900	44123	false	23502301	University of Abomey Calavi	mosquito_habitat_mapper	0.0	2.0	2.0	0.00080	336015	31NDH310051	0.01440	0.01605	0.01580	51262501	0.28655	cement, metal or plastic tank	https://data.globe.gov/system/photos/2024/01/0...	container: artificial	0.000000	identify	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	0.0	0.0	0.249153	-0.242840
2039	POINT (47.50915 -18.90778)	0.0101	0.12430	0.23725	0.20390	0.0939	0.1054	0.1133	0.15750	0.19760	0.2210	0.23230	0.22745	0.23905	true	MDG	Madagascar	GLOBE Observer App	1249.6	LarvaeVisibleNo	[Doris RABENANDRASANA]	0	0.0	NaN	manual	NaN	NaN	NaN	0.00000	0.0	2024-01-03T13:58:00	2024-01-03	0	-18.907100	47.509800	44121	false	52766301	Lycée Andohalo	mosquito_habitat_mapper	0.0	2.0	2.0	0.00050	322792	38KQE643075	0.00960	0.01075	0.01155	23959795	0.24690	discarded: other	https://data.globe.gov/system/photos/2024/01/0...	container: artificial	6.911417	NaN	NaN	NaN	NaN	NaN	NaN	NaN	it rained heavily intermittently yesterday due...	NaN	NaN	NaN	NaN	0.0	0.0	0.344329	-0.375777
2040	POINT (102.78318 17.4057)	0.0085	0.05965	0.23710	0.21785	0.0701	0.0889	0.0864	0.13655	0.19235	0.2066	0.23705	0.22370	0.20380	true	THA	Thailand	GLOBE Observer App	179.2	LarvaeVisibleYes	NaN	307	307.0	NaN	manual	NaN	NaN	NaN	0.00000	0.0	2024-01-03T05:30:00	2024-01-03	0	17.405702	102.783185	44325	false	17615655	Thailand Citizen Science	mosquito_habitat_mapper	0.0	2.0	81.0	0.00050	338811	48QTE645258	0.00715	0.00915	0.00885	113296991	0.27295	well or cistern	https://data.globe.gov/system/photos/2024/01/0...	container: artificial	0.000000	identify-aedes-albopictus	NaN	NaN	4000.0	14693.0	true	false	NaN	NaN	Aedes	albopictus	NaN	0.0	0.0	0.465760	-0.454518
2041	POINT (-89.40154 17.06664)	0.0099	0.25815	0.21230	0.20585	0.2389	0.2491	0.2259	0.25960	0.27315	0.2849	0.28740	0.28865	0.23470	true	GTM	Guatemala	GLOBE Observer App	166.5	LarvaeVisibleYes	NaN	5	5.0	NaN	manual	NaN	NaN	NaN	0.00500	0.0	2024-01-01T13:55:00	2024-01-01	0	17.067000	-89.400800	44119	false	112712735	STEAM Program Guatemala	mosquito_habitat_mapper	0.0	2.0	2.0	0.00090	335850	16QBD444885	0.01600	0.01650	0.01565	113007051	0.34730	old car or boat	NaN	container: artificial	0.000000	identify-no-basal-tuft	NaN	NaN	NaN	NaN	true	false	NaN	https://data.globe.gov/system/photos/2024/01/0...	NaN	NaN	https://data.globe.gov/system/photos/2024/01/0...	0.0	0.0	0.119813	-0.071389

2042 rows × 71 columns

If you ran into any errors with the code above (which may happen due to memory issues), you can run the code below to read the data directly from a copy of it we have uploaded to GitHub.

data = gpd.read_file('https://github.com/geo-di-lab/emerge-lessons/raw/refs/heads/main/docs/data/globe_mosquito_env_2024.zip')

# Remove outliers using the IQR method

larvae_count = data[data['LarvaeCountProcessed'] > 0]

Q1 = larvae_count['LarvaeCountProcessed'].quantile(0.25)
Q3 = larvae_count['LarvaeCountProcessed'].quantile(0.75)
IQR = Q3 - Q1

# identify outliers
threshold = 1.5
data_cleaned = data[(data['LarvaeCountProcessed'] >= Q1 - threshold * IQR) & (data['LarvaeCountProcessed'] <= Q3 + threshold * IQR)]

Plots

fig, (ax1, ax2) = plt.subplots(2)

ax1.hist(data_cleaned[data_cleaned['LarvaeCountProcessed'] > 0]['NDVI'], bins=20)
ax1.set_title("With Larvae")

ax2.hist(data_cleaned[data_cleaned['LarvaeCountProcessed'] == 0]['NDVI'], bins=20)
ax2.set_title("Without Larvae")

plt.suptitle("Distribution of Vegetation (NDVI) at Mosquito Observations")
plt.tight_layout()

fig, (ax1, ax2) = plt.subplots(2)

ax1.hist(data_cleaned[data_cleaned['LarvaeCountProcessed'] > 0]['precipitation'], bins=200)
ax1.set_title("With Larvae")
ax1.set_xlim([1, 20])
ax1.set_ylim([0, 50])

ax2.hist(data_cleaned[data_cleaned['LarvaeCountProcessed'] == 0]['precipitation'], bins=200)
ax2.set_title("Without Larvae")
ax2.set_xlim([1, 20])
ax2.set_ylim([0, 50])

plt.suptitle("Distribution of Precipitation (mm/day) at Mosquito Observations")
plt.tight_layout()

Correlation

from scipy.stats import spearmanr

We will perform Spearman’s Rank Order Correlation using the SciPy Python library. This test gives a correlation coefficient ranging from -1 to 1 measuring the strength of the correlation, where 0 implies no correlation. Positive correlations tell us that as one variable increases, so does the other; and negative correlations tell us that as one decreases, the other decreases as well. Spearman’s Correlation is different from Pearson’s Correlation in that Pearson’s Correlation measures linear relationships, while Spearman’s Correlation can measure non-linear relationships.

cols = ['LarvaeCountProcessed', 'precipitation', 'LST_Day_1km', 'LST_Night_1km', 'NDVI', 'NDWI', 'MeasurementLongitude', 'MeasurementLatitude']
sns.heatmap(data_cleaned[cols].corr(method='spearman'), annot=True, cmap="coolwarm", fmt=".2f", linewidths=0.5)

<Axes: >

def interpret_spearmanr(variable, alpha=0.05):
    """Interprets the statistical significance of a Spearman's correlation result"""
    correlation_coefficient, p_value = spearmanr(data_cleaned[variable], data_cleaned['LarvaeCountProcessed'], nan_policy='omit')

    print(f"Spearman's correlation coefficient: {correlation_coefficient:.3f}")
    print(f"P-value: {p_value}")

    if p_value < alpha:
        direction = "positive" if correlation_coefficient > 0 else "negative"
        print(f"The result is statistically significant, with a {direction} correlation.")
    else:
        print("The result is not statistically significant.")

print(f"Number of observations: {len(data_cleaned)}")

print("\n---NDVI and Mosquito Larvae Count---")
interpret_spearmanr('NDVI')

print("\n---Precipitation and Mosquito Larvae Count---")
interpret_spearmanr('precipitation')

print("\n---Night Temperature and Mosquito Larvae Count---")
interpret_spearmanr('LST_Night_1km')

print("\n---Day Temperature and Mosquito Larvae Count---")
interpret_spearmanr('LST_Day_1km')

print("\n---NDVI and Mosquito Larvae Count---")
interpret_spearmanr('NDVI')

Number of observations: 1500

---NDVI and Mosquito Larvae Count---
Spearman's correlation coefficient: 0.213
P-value: 7.937649861422562e-16
The result is statistically significant, with a positive correlation.

---Precipitation and Mosquito Larvae Count---
Spearman's correlation coefficient: 0.044
P-value: 0.09027661706376768
The result is not statistically significant.

---Night Temperature and Mosquito Larvae Count---
Spearman's correlation coefficient: -0.263
P-value: 1.4908278611951026e-10
The result is statistically significant, with a negative correlation.

---Day Temperature and Mosquito Larvae Count---
Spearman's correlation coefficient: 0.067
P-value: 0.17071659768758848
The result is not statistically significant.

---NDVI and Mosquito Larvae Count---
Spearman's correlation coefficient: 0.213
P-value: 7.937649861422562e-16
The result is statistically significant, with a positive correlation.

Next, we will make plots showing the linear regression model fit using the seaborn Python library. Think of this as drawing a line that tries to fit the relationship between the environmental variables and the count of the mosquito larvae.

fig, (ax1, ax2, ax3) = plt.subplots(1, 3, figsize=(15, 4))

sns.regplot(x = "NDVI", y = "LarvaeCountProcessed", data = data_cleaned[data_cleaned['LarvaeCountProcessed'] > 0],
           line_kws = {"color": "darkblue"}, ax = ax1)
ax1.set_title('Vegetation (NDVI)')

sns.regplot(x = "precipitation", y = "LarvaeCountProcessed", data = data_cleaned[data_cleaned['LarvaeCountProcessed'] > 0],
           line_kws = {"color": "darkblue"}, ax = ax2)
ax2.set_title('Precipitation')

sns.regplot(x = "LST_Night_1km", y = "LarvaeCountProcessed", data = data_cleaned[data_cleaned['LarvaeCountProcessed'] > 0],
           line_kws = {"color": "darkblue"}, ax = ax3)
ax3.set_title('Nighttime Land Surface Temperature')

plt.suptitle('Environmental Conditions at Mosquito Observations')
plt.tight_layout()
plt.show()

References

• SciPy’s spearmanr

• What Is the Interquartile Range (IQR) Rule?

• Get an explanation of statistical significance testing with examples in Python