Summer School

Objectives

1. Classify Vegetation Cover
2. Detect changes in vegetation cover
3. Investigate Vegetation Health

Introduction

The selected study site for monitoring the vegetation was in a protected area located at Marikina watershed in Philippines. The project aims to address vegetation/forest protection and management issues that can be a useful for forest management planning using an Object-Based Image Analysis (OBIA) approach and vegetation indices. For the past 10 years, there has been 4.6% forest cover loss and 6.65% increase in crop land area based on 2003 and 2010 National Mapping and Resource Information Authority (NAMRIA) land cover data. Within the same decade, a 1.90 % annual population growth rate was reported and projected to have 109 million population by the year 2020 based on Philippines Statistics Authority (PSA). Hence, it is important to study the impact of the expansion of urban and agriculture area that initiates deforestation and forest degradation in the long run.

Research Questions

Investigation of vegetation cover change in Marikina watershed area from 1997-2020?

Is there any gain or loss in vegetation cover in Marikina watershed area?

How is the vegetation health assessment changed over the years in the Marikina area?

Methodology

FLOW CHART - CHANGE DETECTION

FLOW CHART - VEGETATION HEALTH

Data Sources

LANDSAT 5 & 8 IMAGERY from USGS earth explorer

Time period: 1997, 2010, 2020

https://earthexplorer.usgs.gov/

SENTINEL-2 DATA from Copernicus Scihub

Time Period: 2020

http ://scihub.copernicus.eu/

Results and Discussion

Vegetation Change

Vegetation Change from 1997-2010

2 types of gain:

small distributed patches within vegetation
large areas bordering the settlements

-> Self-healing / natural reforestation

-> reforestation by government or

Private initiatives or planted specifically to be harvested as timber/charcoal.
Losses come in small distributed patches close to settlements /roads occurred due to local charcoal or timber production

Vegetation Change from 2010-2020

Distributed patches within vegetation fill up, including an area in the southern part of the protected area

-> Self-healing

-> reforestation and recovery after protected area was declared

Loss in distributed patches close to roads, settlements. Areas with previous vegetation gain especially affected !

-> Charcoal / timber making, potential harvesting of previously
planted trees.

Vegetation Health

Conclusion

This study quantified the forest cover loss and gain, landscape pattern changes and forest health assessment in the LMW over the past two decades using OBIA approach.
Challenges and limitations
-> Satellite image acquisition; desired dates/time seasons, cloud-free
-> Threshold identification for vegetation indices; literature search, visual inspection, lack of prior knowledge of study area (vegetation types)
-> Informal/illegal settling, slash and burn farming activities, timber poaching and rapid urbanization leading to severe deforestation and denudation of watershed’s forest cover
Recommendations
->Current watershed rehabilitation program

– establish forest buffer zone outside LMW’s boundary

References

•Santillan J. et al. 2013. Development, calibration and validation of a flood model for Marikina River Basin, Philippines and its applications for flood forecasting, reconstruction and hazard mapping. 10.13140/RG.2.1.3059.2161.

•Clerici et al. 2016. Fusion of Sentinel 1A and Sentinel 2A data for land cover mapping: a case study in the lower Magdalena Region, Colombia.

•Hansen et al. 2001. The use of Sentinel 2 data for mapping European landscapes: the case of Denmark.

OBJECT-BASED CHANGE DETECTION OF VEGETATION IN MARIKINA WATERSHED