TY - JOUR
T1 - Validating modelling and prediction of land use landcover change dynamics based
T2 - 18th International Conference on Quality in Research, QiR 2023
AU - Herdini, Helena Ratih
AU - Gamal, Ahmad
N1 - Publisher Copyright:
© 2024 Author(s).
PY - 2024/11/25
Y1 - 2024/11/25
N2 - This literature study aims to identify the most suitable validation methods for data with a projection period extending beyond the existing maps' temporal coverage. The study employs a comprehensive two-step approach. The initial phase aims to identify patterns and similarities between projected data and established datasets. This study utilizes advanced interpolation and extrapolation techniques, drawing inspiration from previous research. Additionally,the concept of "back casting"is integrated, enabling a retrospective evaluation of the model's accuracy in replicating historical Land Use and Land Cover (LULC) transformations using authentic historical data, aligning with methodologies outlined in the relevant literature. Transitioning to the subsequent phase, the forecasting data spanning the initial decade undergoes a thorough accuracy assessment, sensitivity analysis, and collaboration with stakeholders, culminating in refining the model following procedures delineated in existing scholarly works. Once the model is calibrated, it predicts the trajectory of LULC changes for the ensuing thirty years, incorporating various scenarios and quantifying uncertainties in line with recommendations from relevant studies. The transparent dissemination of resultsand a comprehensive evaluation across diverse study regions collectively establish the credibility and applicability of the study's findings regarding historical LULC changes as a guiding framework for shaping future predictive models.
AB - This literature study aims to identify the most suitable validation methods for data with a projection period extending beyond the existing maps' temporal coverage. The study employs a comprehensive two-step approach. The initial phase aims to identify patterns and similarities between projected data and established datasets. This study utilizes advanced interpolation and extrapolation techniques, drawing inspiration from previous research. Additionally,the concept of "back casting"is integrated, enabling a retrospective evaluation of the model's accuracy in replicating historical Land Use and Land Cover (LULC) transformations using authentic historical data, aligning with methodologies outlined in the relevant literature. Transitioning to the subsequent phase, the forecasting data spanning the initial decade undergoes a thorough accuracy assessment, sensitivity analysis, and collaboration with stakeholders, culminating in refining the model following procedures delineated in existing scholarly works. Once the model is calibrated, it predicts the trajectory of LULC changes for the ensuing thirty years, incorporating various scenarios and quantifying uncertainties in line with recommendations from relevant studies. The transparent dissemination of resultsand a comprehensive evaluation across diverse study regions collectively establish the credibility and applicability of the study's findings regarding historical LULC changes as a guiding framework for shaping future predictive models.
UR - http://www.scopus.com/inward/record.url?scp=85212227723&partnerID=8YFLogxK
U2 - 10.1063/5.0235702
DO - 10.1063/5.0235702
M3 - Conference article
AN - SCOPUS:85212227723
SN - 0094-243X
VL - 3215
JO - AIP Conference Proceedings
JF - AIP Conference Proceedings
IS - 1
M1 - 030004
Y2 - 23 October 2023 through 25 October 2023
ER -