Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Investigating the use of Landsat TM for mapping leaf nitrogen of Norway spruce

Lindquist, Stefan (2002) In Lunds universitets Naturgeografiska institution - Seminarieuppsatser
Dept of Physical Geography and Ecosystem Science
Abstract (Swedish)
Populärvetenskaplig sammanfattning: Kvävekoncentrationen i den gröna biomassan hos vegetation är involverad i många olika
biokemiska processer i det terrestra ekosystemet. Till exempel så påverkar den
fotosyntesen, förmultning och evaportanspirationen, för att nämna några. Därför är den
också ofta en av grundparametrarna och drivkraften i olika ekosystem-modeller. Att mäta
kvävehalten hos blad/barr i fält kan vara både tidsödande och dyrt, varför det vore
värdefullt att kunna uppskatta den med hjälp av fjärranalys. I olika studier har man
använt: statistiska-, empiriska/matematiska metoder, eller lite mer sofistikerade
reflektansmodeller. Alla vilka ämnar uppskatta eller hitta samband mellan reflektans och
kvävekoncentrationen i... (More)
Populärvetenskaplig sammanfattning: Kvävekoncentrationen i den gröna biomassan hos vegetation är involverad i många olika
biokemiska processer i det terrestra ekosystemet. Till exempel så påverkar den
fotosyntesen, förmultning och evaportanspirationen, för att nämna några. Därför är den
också ofta en av grundparametrarna och drivkraften i olika ekosystem-modeller. Att mäta
kvävehalten hos blad/barr i fält kan vara både tidsödande och dyrt, varför det vore
värdefullt att kunna uppskatta den med hjälp av fjärranalys. I olika studier har man
använt: statistiska-, empiriska/matematiska metoder, eller lite mer sofistikerade
reflektansmodeller. Alla vilka ämnar uppskatta eller hitta samband mellan reflektans och
kvävekoncentrationen i löv. De flesta av studierna har den gemensamma nämnaren att de
använt sig av hög upplösande reflektansdata, eller så kallad hyper-spektral data.
I denna studie har en undersökning utförts som utvärderar hur lämplig den spektrala
information man kan tillhandahålla från Landsat TM är. Och ifall dess reflektansvärden
har någon koppling till kvävehalten i barren hos gran (Picea abis (L.) Karst.). Utöver det
har även en utvärdering av reflektansmodellen PROSPECT gjorts, både för att få veta
vilka parametrar som är viktigast och för att se hur de påverkar den modellerade
reflektansen. Ett fältarbeta utfördes på våren år 2000 där tio stycken granskogsbestånd
besöktes i fält. Olika beståndsparametrar mättes, däribland kvävehalten hos barren.
Ingen korrelation kunde fastställas mellan reflektansdatan och kvävehalten i barren, den
högsta korrelationskoefficienten var 0.48 (N.S). Däremot så gjordes även en stepwise
regression där två TM kanaler valdes ut, TM1 och TM2, detta med ett R2-värde på 0.72.
Detta samband, baserat på andra studier, anses komma från korrelation mellan kväve och
klorofyll och inte det direkta förhållandet mellan reflektans och kväve i barren.
Utvärderingen av PROSPECT modellen visade att kvävehalten hade väldigt litet
inflytande på den simulerade reflektansen. Det största inflytande hade antalet aktiva
cellager i lövet och dess klorofyllkoncentration. En gradient från väster till öster i bladets kvävehalt har upptäckts, detta med hjälp av
regressionsanalys mellan X-koordinat och kvävekoncentration (R2 = 0.77). Gradienten
kunde inte observeras i fjärranalysdatan. (Less)
Abstract
Nitrogen in the green biomass of vegetation plays an important role in many biochemical
processes in the terrestrial ecosystems, and often used as an input parameter to ecosystem
models. It strongly influences photosynthetic rate, decomposition, and
evaportanspiration, to name a few. Measuring nitrogen in field involves a lot of time and
money, and it would be valuable if this part of the data collection could be done with
remotely sensed data. In recent research, different methods has been used, statistical
relationship, empirical/mathematical approaches or more sophisticated with the use of
advance reflectance modeling, that all attempts to estimate or find a connection to
nitrogen through remotely sensed data. Most of them have... (More)
Nitrogen in the green biomass of vegetation plays an important role in many biochemical
processes in the terrestrial ecosystems, and often used as an input parameter to ecosystem
models. It strongly influences photosynthetic rate, decomposition, and
evaportanspiration, to name a few. Measuring nitrogen in field involves a lot of time and
money, and it would be valuable if this part of the data collection could be done with
remotely sensed data. In recent research, different methods has been used, statistical
relationship, empirical/mathematical approaches or more sophisticated with the use of
advance reflectance modeling, that all attempts to estimate or find a connection to
nitrogen through remotely sensed data. Most of them have the thing in common that they
use high-resolution reflectance data, so called hyper-spectral data.
In this study an investigation was done to find out how useful the spectral information
from Landsat TM is, and if the reflectance values has any connection the needle nitrogen
concentration of Norway spruce (Picea abis (L.) Karst.). Moreover, an evaluation of the
PROSPECT model was done, both to find out the importance of each input parameter
and to find out how they affect the reflectance. A fieldwork was carried out in the spring
of 2000 where ten different Norway spruce stands in Scania were visited and various
parameters measured, including the concentration of nitrogen in the needles.
No correlation was found between the reflectance data and nitrogen in the needles, the
highest correlation coefficient was 0.48 (N.S). However, in a stepwise regression two TM
channels were selected, TM1 and TM2, with an R2-value of 0.72. This relationship, based
on other studies, is assumed to origin from the correlation between nitrogen and
chlorophyll and not the direct relation between reflectance and the needle nitrogen.
The evaluation of the PROSPECT model showed that nitrogen concentrations in leaves
have little effect on the simulated reflectance. The biggest influence on the reflectance
seems to be the active number of cell layers in the leaf and the leaves’ chlorophyll
concentration. A gradient from west to east in the leaves nitrogen concentration was found, this with a
regression analysis between easting coordinate (X) and nitrogen concentration (R2 =
0.77). The gradient could not be observed in the remotely sensed data. (Less)
Please use this url to cite or link to this publication:
author
Lindquist, Stefan
supervisor
organization
year
type
H1 - Master's Degree (One Year)
subject
keywords
geomorfologi, naturgeografi, climatology, remote sensing, modelling, physical geography, reflectance, geomorphology, cartography, pedology, marklära, kartografi, klimatologi
publication/series
Lunds universitets Naturgeografiska institution - Seminarieuppsatser
report number
92
language
English
id
1333051
date added to LUP
2005-10-24 00:00:00
date last changed
2011-12-07 10:32:33
@misc{1333051,
  abstract     = {{Nitrogen in the green biomass of vegetation plays an important role in many biochemical
processes in the terrestrial ecosystems, and often used as an input parameter to ecosystem
models. It strongly influences photosynthetic rate, decomposition, and
evaportanspiration, to name a few. Measuring nitrogen in field involves a lot of time and
money, and it would be valuable if this part of the data collection could be done with
remotely sensed data. In recent research, different methods has been used, statistical
relationship, empirical/mathematical approaches or more sophisticated with the use of
advance reflectance modeling, that all attempts to estimate or find a connection to
nitrogen through remotely sensed data. Most of them have the thing in common that they
use high-resolution reflectance data, so called hyper-spectral data.
In this study an investigation was done to find out how useful the spectral information
from Landsat TM is, and if the reflectance values has any connection the needle nitrogen
concentration of Norway spruce (Picea abis (L.) Karst.). Moreover, an evaluation of the
PROSPECT model was done, both to find out the importance of each input parameter
and to find out how they affect the reflectance. A fieldwork was carried out in the spring
of 2000 where ten different Norway spruce stands in Scania were visited and various
parameters measured, including the concentration of nitrogen in the needles.
No correlation was found between the reflectance data and nitrogen in the needles, the
highest correlation coefficient was 0.48 (N.S). However, in a stepwise regression two TM
channels were selected, TM1 and TM2, with an R2-value of 0.72. This relationship, based
on other studies, is assumed to origin from the correlation between nitrogen and
chlorophyll and not the direct relation between reflectance and the needle nitrogen.
The evaluation of the PROSPECT model showed that nitrogen concentrations in leaves
have little effect on the simulated reflectance. The biggest influence on the reflectance
seems to be the active number of cell layers in the leaf and the leaves’ chlorophyll
concentration. A gradient from west to east in the leaves nitrogen concentration was found, this with a
regression analysis between easting coordinate (X) and nitrogen concentration (R2 =
0.77). The gradient could not be observed in the remotely sensed data.}},
  author       = {{Lindquist, Stefan}},
  language     = {{eng}},
  note         = {{Student Paper}},
  series       = {{Lunds universitets Naturgeografiska institution - Seminarieuppsatser}},
  title        = {{Investigating the use of Landsat TM for mapping leaf nitrogen of Norway spruce}},
  year         = {{2002}},
}