MINI PROJECT

​

By

​

​

Rithik Akash.M (2017107046)

Mukil.M (2017107038)

Arthi.K (2017107055)

​

PROJECT TITLE

Identifying Oak Tree Mortality in Himalayas

Using Image Analysis

​

​

​

​

​

CONTENTS

• Introduction
• Causes for Oak tree depletion
• Objectives
• Concept and background literature
• Study Area
• Data Products and Softwares used
• Satellite Image Properties
• Implementation
• Classification
• Observation
• Results
• Conclusion
• References

​

Introduction

• Plant species like oak tree are in an endangered situation which needs care and attention to preserve. They have been applying different methods to treat and restore the environment at different time and place in a haphazard manner as indicated by the customer.
• The present biological system is exceedingly convoluted and fluctuated which requires close checking to measure change and react successfully and proficiently with different procedures to deal with the abatement in plant biodiversity.

Causes for Oak tree depletion

• Oaks are the 'people's species', providing fodder for live-stock, firewood for cooking, tools for agricultural activity, and leaves for manuring the cropfields. Their presence has been related to the quality and quantity of spring water. In other words, oak forests have been the basis of age-old subsistence agriculture in these mountains.
• Oaks are the 'people's species', providing fodder for live-stock, firewood for cooking, tools for agricultural activity, and leaves for manuring the cropfields. Their presence has been related to the quality and quantity of spring water. In other words, oak forests have been the basis of age-old subsistence agriculture in these mountains.

Objective

1. To develop a customized GIS tool and its Python script for determining mortality locations and trends.
2. To deliver a file geodatabase with all documented data and information.
3. To deliver a map showing tree mortality trends in the targeted areas.

Concept and Background literature

Normalized Difference Vegetation Index (NDVI) parameter and its importance for identifying healthy trees are used in this project.

Satellite images have a capability to distinguish regions containing dead trees, yet there was an issue of an excess mixing of spectral signatures with other land cover types. Vegetation indices alone were insufficient to distinguish dead trees and scrub due to spectral signature overlap. To correctly identify dead trees using satellite imagery, the supportive landcover datasets that recognize non-vegetated regions and the zones with a known reason for death were necessary.

​

Study Area

Oaks are an anomaly in a largely tropical country like India, but they form the backbone of the temperate moist forest biome occurring between 1,500 m and 2,700 m in the Himalayas. Five oak species are found in the Western Himalaya of which the most widespread one is the banj oak or Quercus leucotrichophora, that occurs as the dominant species up to about 2,200 m. 

​

Data Products and softwares used

• Satellite imagery from USGS Earth Explorer
1. (LANDSAT - 8)
• Esri image services and other GIS data layers like base maps.
• ArcGIS 10.8, Google Earth Pro are applicable and are used for the project.

Satellite Image Properties

• Downloaded from : LANDSAT 8
• Source : USGS Earth Explorer
• Location :Srinagar
• PATH : 149
• ROW : 36
• Level of Classification : Level -2, Classification – 2
• Observation time : July 8, 2020
• Chosen bands : Band 4 and Band 5

​

Band 4 image

Band 5 image

Implementation

1. Factors Controlling Leaf Reflectance

The reflectance and transmittance properties of leaves are influenced by leaf colors, internal structure, and leaf water content.

The photosynthetic process is:

6CO₂ + 6H₂O + light energy = C₆H₁₂O ₆ + 6O₂

2. Near-Infrared Energy Interaction with Plant Cells

A healthy green leaves near-infrared reflectance increases in a range from 700-1200 nm. In contrast, healthy green leaves absorb the visible part of the range where incoming light is essential for photosynthesis.

NDVI Image

NDVI values and plant health

Classification

Based on the relationship between NDVI values and plant health, the satellite image is categories as the following:

• Dead plant
• Unhealthy plant
• Moderately healthy plant
• Very healthy plant

​

Observation

• Temporal resolution also impacts NDVI values. The amount of reflected and absorbed electromagnetic energy by particles shifts in time, and vegetations have an alternate measure of chlorophyll in various seasons.
• Atmospheric particles disperse electromagnetic energy toward all path. The red portion of electromagnetic energy has more dissipated by environmental particles than near-infrared energy. This has an impact of diminishing NDVI values.
• Soil type especially soil moisture, have a significant effect on NDVI values, a probability of up to 20 percent change in NDVI result. Soil with greater radiance reflected have lower NDVI values even in similar vegetative conditions.

Results

Conclusion

• The study area does not contain very healthy trees. The percentage of trees that are moderately healthy is only 60%, unhealthy plants is 30% and that of dead trees is 10%. From this it is evident that the number of trees is depleting and the health status is poor.
• Living and nonliving variables are disturbing the sound development of a tree.
• Expanding the information of oak tree mortality utilizing remote sensing innovation would support preservationists, researchers, naturalists, and supervisors to avert further mortality and take important treatment measures.

​

References

• Identifying Oak Tree Mortality Using Image Analysis (Zemenu Teshome Ambelu - University of Redlands)
• Modeling individual-tree mortality in Pyrenean oak (Quercus pyrenaica Willd.)
• Oak mortality risk factors and mortality estimation Stephen R. Shifley , Zhaofei Fan , John M. Kabrick , Randy G. Jensen

​

​

Thank You!