City University of Hong Kong

CityU Institutional Repository >
3_CityU Electronic Theses and Dissertations >
ETD - Dept. of Physics and Materials Science  >
AP - Master of Philosophy  >

Please use this identifier to cite or link to this item:

Title: An examination on the effectiveness of extended Kalman filter in land surface data assimilation
Other Titles: Dui Kamen lu bo zai lu mian shu zhi tong hua zhong you xiao xing de tan tao
Authors: Leung, Ka Yan (梁嘉恩)
Department: Dept. of Physics and Materials Science
Degree: Master of Philosophy
Issue Date: 2006
Publisher: City University of Hong Kong
Subjects: Atmospheric radiation -- Mathematical models
Kalman filtering
Notes: CityU Call Number: QC912.3.L48 2006
Includes bibliographical references (leaves 89-91)
Thesis (M.Phil.)--City University of Hong Kong, 2006
v, 91 leaves : ill. (some col.) ; 30 cm.
Type: Thesis
Abstract: Land surface is a key component of the climate system. The exchanges of energy, mass and momentum between the land surface and the atmosphere take place on a variety of spatial and temporal scales. For the representation of land surface processes in atmospheric and surface hydrologic models, a number of land surface schemes have been developed. However, land surface schemes have inevitably large uncertainties because land surface processes are complex and land surface parameters vary strongly in space and time. The accuracy of land surface modeling can be improved through the assimilation of observated data. In this study, we investigate the effectiveness of an extended Kalman filter (EKF) in the assimilation of surface radiative temperature into land surface modeling. The central problems a land surface scheme must deal with are surface energy and water balances. These two balances are coupled through evaporation. Different land surface schemes have different configurations, ranging from singlesoil- layer to multi-soil-layer. The land surface scheme used for this study, ALSIS (Atmospheric-Land-Surface Interaction Scheme), is a multi-soil-layer model. ALSIS is used to simulate surface energy fluxes, evapotranspiration and surface soil hydrological quantities, such as runoff and drainage. This study consists of two components: an examination on the performance of ALSIS and an examination of the effectiveness of the extended Kalman filter for land surface data assimilation. For the first component, ALSIS is run offline using the data from a field experiment carried out at Qira. The simulated results are compared with the field measurements. For the second component, the HAPEXMOBILHY data are used. All models of physical and other processes do not perfectly describe the reality. They are only approximations. A useful method for improving the model predictions is to make use of the actual observations to nudge the model simulations close to the reality. This technique, known as data assimilation, has been widely applied to atmosphere and ocean modeling, but has so far not been widely used for land surface modeling. In this work, two data assimilation schemes, the EKF and the forced correction method, are tested in the framework of ALSIS. The EKF theory is first described. A well-calibrated simulation of ALSIS is used as the reference case, i.e., the truth. Based on the “true” radiative and canopy temperatures, the microwave brightness temperature of the surface is estimated. The EKF is then used to assimilate the brightness temperature to ALSIS simulations using data with various contaminations. In case of the forced correction scheme, the model radiative temperature is replaced by the “true” radiative temperature at each time step. The moisture of the first soil layer is studied to check the assimilation performance. To test the assimilation ability, pre-specified errors are introduced to the atmospheric forcing data (Experiment A) and the input land surface parameters (Experiment B). For Experiment A, the false data include precipitation, long wave radiation and short wave radiation. The error contaminated input precipitation is assumed to be twice the true precipitation and the error contaminated long wave and short wave radiations are the true values superposed with Gaussian noises. For Experiment B, the standard soil type is loam and the test soil type is sandy loam and sandy clay loam. The erroneous input of the roughness length is 0.05 m which is 5 times the original roughness length 0.01 m. Both the EKF and the forced correction methods are found to work well under certain conditions. For some cases, the soil moisture simulated with the EKF reaches unrealistic value. The reasons for these extreme values are examined by studying the Kalman filter parameters, in particular the Kalman gain and the error covariance propagation. The overall assimilation performance over the whole period of time is determined by the degree of agreement and the correlation coefficient.
Online Catalog Link:
Appears in Collections:AP - Master of Philosophy

Files in This Item:

File Description SizeFormat
fulltext.html159 BHTMLView/Open
abstract.html159 BHTMLView/Open

Items in CityU IR are protected by copyright, with all rights reserved, unless otherwise indicated.


Valid XHTML 1.0!
DSpace Software © 2013 CityU Library - Send feedback to Library Systems
Privacy Policy · Copyright · Disclaimer