1. Syed Mohammad Sualeh Ali
Muhammed Maaz Tariq
Talat Zubair
Supervisor:
Dr. Tariq Mehmood
2. Product Overview
This project is part of the Environment Monitoring
module of NED RCAI Lab’s CitiPulse Mega Portal
project, that is working towards urban data
visualization and analysis.
ShajrStats is an interactive web application that
provides quantifiable information about the levels
of vegetation throughout Pakistan, making it
possible to measure and compare the amount of
‘Green Areas’ in different regions.
3. Problem Statement
We do not have any tools in Pakistan that measure or keep track of vegetation
growth or depletion. Even if the records are available, they are unconsolidated
and insufficient for a detailed analysis. Due to which:
- We do not have a way to identify areas needed for plantation,
- There is a lack of Urban Planning in government initiatives,
- No environmental Monitoring,
- Students and researchers looking to advance their analyses in this domain
find their progress hampered,
- No transparency in audit for green initiatives.
4. Use Cases
➔ Urban Planning
➔ Plantation Drives
➔ Environmental Research
➔ Health-Related Research
➔ Real-Estate Businesses
Can be used to maintain certain level of greenery
within each town.
Can be used to identify which areas have low
amount of green area and need more plantations.
Can be used to identify properties that have parks,
grounds etc nearby.
Can be used as an independent variable to study
the impacts on Temperature, Rainfall, Humidity etc.
Can be used to study the impact of greenery on
public health.
5. Road Map
We used Google Earth
Engine and obtained the
data from its catalog of
Landsat imagery. The
analysis was done using
the normalized difference
vegetation index (ndvi)
One of the major issues
we faced was getting the
satellite data. We were
unable to find freely
available high-resolution
satellite imagery for
Pakistan that could be
used for analysis.
We used the Planet API
to extract imagery from
Planet.com, but the data
quality was poor in the
free version and the
premium had exorbitant
pricing schemes.
Initial Stage (Obtaining Data) First Approach to Implementation Second Approach to Implementation
6. Our Approach
● Imported Landsat Satellite Imagery in Google Earth Engine
● Defined the coordinates for regional boundaries using GeoJSON
● Set a Threshold value for NDVI and Cloud Cover
● Performed a Binary Classification by classifying all values greater than the
Threshold as 1, and the rest as 0.
● Used a Reducer Function to sum up the amount of green area
● Used Heroku for app deployment
8. Future Work
● Obtaining the data for soil quality of different regions can further help
determine what type of vegetation needs to be cultivated in which areas.
● Adding the air quality index for each region can be used to quantitatively
study the correlation between air pollution and deforestation.
● Over-time temperature metrics can help correlate plantation activities
with climate change.
● Real-time analysis of Greenery KPIs