Polynomial trends of vegetation phenology in Sahelian to equatorial Africa using remotely sensed time series from 1983 to 2005
(2014) In Student thesis series INES NGEM01 20122Dept of Physical Geography and Ecosystem Science
- Abstract
- Popular science
Our understanding of global warming can be achieved in different ways. One way is to study the phenological parameters of vegetation. Phenology or seasonality of vegetation can be identified from several parameters such as: the start of the growing season (SOS), end of the growing season (EOS), amplitude of the season (AMP), and length of the growing season (LOS). Changes of these parameters represent the cyclic changes of vegetation. Nowadays, imagery satellite data are reliable and widely-used sources to study the vegetation changes. Phenology parameters are derived from time series of vegetation indices (VI) that can be computed from satellite imagery. In this thesis, long-term dataset of GIMMS NDVI from 1983 to 2005... (More) - Popular science
Our understanding of global warming can be achieved in different ways. One way is to study the phenological parameters of vegetation. Phenology or seasonality of vegetation can be identified from several parameters such as: the start of the growing season (SOS), end of the growing season (EOS), amplitude of the season (AMP), and length of the growing season (LOS). Changes of these parameters represent the cyclic changes of vegetation. Nowadays, imagery satellite data are reliable and widely-used sources to study the vegetation changes. Phenology parameters are derived from time series of vegetation indices (VI) that can be computed from satellite imagery. In this thesis, long-term dataset of GIMMS NDVI from 1983 to 2005 was used to extract and analyze vegetation phenology over Sahelian to equatorial areas. The TIMESAT software package was also used as an automated method to extract the parameters.
Recent researches have shown the changes via analyzing the linear trends of the vegetation indices or lately through studying the linear trend of phenological parameters. Since changes of vegetation are not always simply linear, the overall aim of this thesis was to study vegetation changes through analysis of non-linear trends and more complex mathematical functions of phenology parameters, and via finding the relationship between the phenology parameters and soil moisture. Driving forces behind changes in phenology parameters including land cover, soil texture and rainfall were also taken in to consideration.
The results illustrated that non-linear trends can detect notable proportions of vegetation changes in the study area. Not only significant portions of areas with linear trends could be represented using non-linear trends, but also these trends increased the precision of phenology change detection. Regarding the climate driver forces results showed that the vegetation phenology changes followed soil moisture variations. However the trends of vegetation changes has not especially followed land cover, soil texture and geographic characteristics although in some limited cases these driver forces are related to the changes. (Less) - Abstract
- Global warming has both short and long term effects on seasonal phenological cycles of vegetation. Phenology parameters of vegetation such as start, end, length and amplitude of season can describe life cycle events of vegetation. In this thesis, long-term dataset of GIMMS NDVI time series from 1983 to 2005 was used to extract and analyze vegetation phenology over Sahelian to equatorial areas and TIMESAT software package was used as an automated method to extract the parameters.
The overall aim of this thesis was to study vegetation changes through analysis of polynomial trends of phenology parameters. Phenology parameters were analyzed to detect hidden changes in vegetation dynamics. Through comparing polynomial trends of vegetation... (More) - Global warming has both short and long term effects on seasonal phenological cycles of vegetation. Phenology parameters of vegetation such as start, end, length and amplitude of season can describe life cycle events of vegetation. In this thesis, long-term dataset of GIMMS NDVI time series from 1983 to 2005 was used to extract and analyze vegetation phenology over Sahelian to equatorial areas and TIMESAT software package was used as an automated method to extract the parameters.
The overall aim of this thesis was to study vegetation changes through analysis of polynomial trends of phenology parameters. Phenology parameters were analyzed to detect hidden changes in vegetation dynamics. Through comparing polynomial trends of vegetation parameters and soil moisture, the relationship between the phenology parameters and soil moisture was detected and the role of climate driver forces (including land cover, soil texture and rainfall) behind the changes in phenology parameters were investigated.
The results illustrated that polynomial trends can detect notable proportions of vegetation changes in the Sahel using remotely sensed data. Significant portions of areas with linear trends could be represented through quadratic and cubic trends, and these trends increased the precision of phenology change detection. Furthermore, in some areas vegetation changes were not detected neither through linear regressions nor polynomial trends. In such areas, polynomial hidden trends could be applied for detecting the fluctuations of vegetation parameters. In summation, applying polynomial trend analysis to time-series of satellite data is a powerful tool for investigating trends and variations in vegetation in semi-arid to sub-humid regions, like the Sahel.
Regarding the climate driver forces, results showed that the vegetation phenology changes followed soil moisture variations, and in most occurrences, moderate correlations were found between SOS, EOS, and soil moisture. The trends of vegetation changes did not spatially follow land cover and soil types of the study area. However, in some limited cases, land cover, soil texture and geographic characteristics such as elevation were related to the changes. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/4387849
- author
- Veysipanah, Mozafar LU
- supervisor
-
- Lars Eklundh LU
- organization
- course
- NGEM01 20122
- year
- 2014
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- TIMESAT, remote sensing, GIMMS NDVI, Physical Geography and Ecosystem analysis, Sahel, phenology parameters, polynomial trends, vegetation, soil moisture
- publication/series
- Student thesis series INES
- report number
- 304
- language
- English
- id
- 4387849
- date added to LUP
- 2014-04-02 16:39:40
- date last changed
- 2014-04-02 16:39:40
@misc{4387849,
abstract = {{Global warming has both short and long term effects on seasonal phenological cycles of vegetation. Phenology parameters of vegetation such as start, end, length and amplitude of season can describe life cycle events of vegetation. In this thesis, long-term dataset of GIMMS NDVI time series from 1983 to 2005 was used to extract and analyze vegetation phenology over Sahelian to equatorial areas and TIMESAT software package was used as an automated method to extract the parameters.
The overall aim of this thesis was to study vegetation changes through analysis of polynomial trends of phenology parameters. Phenology parameters were analyzed to detect hidden changes in vegetation dynamics. Through comparing polynomial trends of vegetation parameters and soil moisture, the relationship between the phenology parameters and soil moisture was detected and the role of climate driver forces (including land cover, soil texture and rainfall) behind the changes in phenology parameters were investigated.
The results illustrated that polynomial trends can detect notable proportions of vegetation changes in the Sahel using remotely sensed data. Significant portions of areas with linear trends could be represented through quadratic and cubic trends, and these trends increased the precision of phenology change detection. Furthermore, in some areas vegetation changes were not detected neither through linear regressions nor polynomial trends. In such areas, polynomial hidden trends could be applied for detecting the fluctuations of vegetation parameters. In summation, applying polynomial trend analysis to time-series of satellite data is a powerful tool for investigating trends and variations in vegetation in semi-arid to sub-humid regions, like the Sahel.
Regarding the climate driver forces, results showed that the vegetation phenology changes followed soil moisture variations, and in most occurrences, moderate correlations were found between SOS, EOS, and soil moisture. The trends of vegetation changes did not spatially follow land cover and soil types of the study area. However, in some limited cases, land cover, soil texture and geographic characteristics such as elevation were related to the changes.}},
author = {{Veysipanah, Mozafar}},
language = {{eng}},
note = {{Student Paper}},
series = {{Student thesis series INES}},
title = {{Polynomial trends of vegetation phenology in Sahelian to equatorial Africa using remotely sensed time series from 1983 to 2005}},
year = {{2014}},
}