A new approach to LST modeling and normalization under clear-sky conditions based on a local optimization strategy
The normalization of LST relative to environmental parameters is of great importance in various environmental applications. The purpose of this study was to develop a new approach for LST normalization relative to environmental variables. These included topographic variables (i.e. solar irradiance a...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Taylor & Francis Group
2022-12-01
|
Series: | International Journal of Digital Earth |
Subjects: | |
Online Access: | http://dx.doi.org/10.1080/17538947.2022.2137254 |
_version_ | 1797678467347644416 |
---|---|
author | Majid Kiavarz Mohammad Karimi Firozjaei Seyed Kazem Alavipanah Quazi K. Hassan Yoann Malbéteau Si-Bo Duan |
author_facet | Majid Kiavarz Mohammad Karimi Firozjaei Seyed Kazem Alavipanah Quazi K. Hassan Yoann Malbéteau Si-Bo Duan |
author_sort | Majid Kiavarz |
collection | DOAJ |
description | The normalization of LST relative to environmental parameters is of great importance in various environmental applications. The purpose of this study was to develop a new approach for LST normalization relative to environmental variables. These included topographic variables (i.e. solar irradiance and near-surface temperature lapse rate (NSTLR)) as well as biophysical properties (i.e. vegetation, wetness, and albedo). The study was conducted in two phases, namely (1) using global and (2) local optimization strategies to calculate the regression coefficients of environmental variables in the partial least squares regression (PLSR) and build the non-linear linking model in the random forest regression (RFR). The RMSEs between actual LST and modeled LST based on the global and local optimization strategies using PLSR (RFR) were 2.202 (0.935) and 0.939 (0.835) °C, respectively. The results showed that RFR had higher efficiency than PLSR in normalizing LST. Moreover, the local optimization method outperformed the global optimization method in terms of normalization accuracy. The results of this study could be very useful in many environmental applications such as identifying thermal anomalies, and surface anthropogenic heat island modeling. |
first_indexed | 2024-03-11T23:00:11Z |
format | Article |
id | doaj.art-0ebe29971f304ecc87d5dbea6211826c |
institution | Directory Open Access Journal |
issn | 1753-8947 1753-8955 |
language | English |
last_indexed | 2024-03-11T23:00:11Z |
publishDate | 2022-12-01 |
publisher | Taylor & Francis Group |
record_format | Article |
series | International Journal of Digital Earth |
spelling | doaj.art-0ebe29971f304ecc87d5dbea6211826c2023-09-21T14:57:11ZengTaylor & Francis GroupInternational Journal of Digital Earth1753-89471753-89552022-12-011511833185410.1080/17538947.2022.21372542137254A new approach to LST modeling and normalization under clear-sky conditions based on a local optimization strategyMajid Kiavarz0Mohammad Karimi Firozjaei1Seyed Kazem Alavipanah2Quazi K. Hassan3Yoann Malbéteau4Si-Bo Duan5University of TehranUniversity of TehranUniversity of TehranUniversity of CalgaryMohammed VI Polytechnic University (UM6P)Ministry of Agriculture/Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural SciencesThe normalization of LST relative to environmental parameters is of great importance in various environmental applications. The purpose of this study was to develop a new approach for LST normalization relative to environmental variables. These included topographic variables (i.e. solar irradiance and near-surface temperature lapse rate (NSTLR)) as well as biophysical properties (i.e. vegetation, wetness, and albedo). The study was conducted in two phases, namely (1) using global and (2) local optimization strategies to calculate the regression coefficients of environmental variables in the partial least squares regression (PLSR) and build the non-linear linking model in the random forest regression (RFR). The RMSEs between actual LST and modeled LST based on the global and local optimization strategies using PLSR (RFR) were 2.202 (0.935) and 0.939 (0.835) °C, respectively. The results showed that RFR had higher efficiency than PLSR in normalizing LST. Moreover, the local optimization method outperformed the global optimization method in terms of normalization accuracy. The results of this study could be very useful in many environmental applications such as identifying thermal anomalies, and surface anthropogenic heat island modeling.http://dx.doi.org/10.1080/17538947.2022.2137254normalizationland surface temperatureenvironmental variablespartial least squaresrandom forestlocal strategy |
spellingShingle | Majid Kiavarz Mohammad Karimi Firozjaei Seyed Kazem Alavipanah Quazi K. Hassan Yoann Malbéteau Si-Bo Duan A new approach to LST modeling and normalization under clear-sky conditions based on a local optimization strategy International Journal of Digital Earth normalization land surface temperature environmental variables partial least squares random forest local strategy |
title | A new approach to LST modeling and normalization under clear-sky conditions based on a local optimization strategy |
title_full | A new approach to LST modeling and normalization under clear-sky conditions based on a local optimization strategy |
title_fullStr | A new approach to LST modeling and normalization under clear-sky conditions based on a local optimization strategy |
title_full_unstemmed | A new approach to LST modeling and normalization under clear-sky conditions based on a local optimization strategy |
title_short | A new approach to LST modeling and normalization under clear-sky conditions based on a local optimization strategy |
title_sort | new approach to lst modeling and normalization under clear sky conditions based on a local optimization strategy |
topic | normalization land surface temperature environmental variables partial least squares random forest local strategy |
url | http://dx.doi.org/10.1080/17538947.2022.2137254 |
work_keys_str_mv | AT majidkiavarz anewapproachtolstmodelingandnormalizationunderclearskyconditionsbasedonalocaloptimizationstrategy AT mohammadkarimifirozjaei anewapproachtolstmodelingandnormalizationunderclearskyconditionsbasedonalocaloptimizationstrategy AT seyedkazemalavipanah anewapproachtolstmodelingandnormalizationunderclearskyconditionsbasedonalocaloptimizationstrategy AT quazikhassan anewapproachtolstmodelingandnormalizationunderclearskyconditionsbasedonalocaloptimizationstrategy AT yoannmalbeteau anewapproachtolstmodelingandnormalizationunderclearskyconditionsbasedonalocaloptimizationstrategy AT siboduan anewapproachtolstmodelingandnormalizationunderclearskyconditionsbasedonalocaloptimizationstrategy AT majidkiavarz newapproachtolstmodelingandnormalizationunderclearskyconditionsbasedonalocaloptimizationstrategy AT mohammadkarimifirozjaei newapproachtolstmodelingandnormalizationunderclearskyconditionsbasedonalocaloptimizationstrategy AT seyedkazemalavipanah newapproachtolstmodelingandnormalizationunderclearskyconditionsbasedonalocaloptimizationstrategy AT quazikhassan newapproachtolstmodelingandnormalizationunderclearskyconditionsbasedonalocaloptimizationstrategy AT yoannmalbeteau newapproachtolstmodelingandnormalizationunderclearskyconditionsbasedonalocaloptimizationstrategy AT siboduan newapproachtolstmodelingandnormalizationunderclearskyconditionsbasedonalocaloptimizationstrategy |