Assessment of Land-Use Scenarios at a National Scale Using Intensity Analysis and Figure of Merit Components

Abstract

To address the impacts of future land changes on biodiversity and ecosystem services, land-use scenarios have been developed at the national scale in Japan. However, the validation of land-use scenarios remains a challenge owing to the lack of an appropriate validation method. This research developed land-use maps for 10 land-use categories to calibrate a land-change model for the 1987–1998 period, simulate changes during the 1998–2014 period, and validate the simulation for the 1998–2014 period. Following an established method, this study assessed the three types of land change: (1) reference change during the calibration time interval, (2) simulation change during the validation time interval, and (3) reference change during the validation time interval, using intensity analysis and figure of merit components (hits, misses, and false alarms). The results revealed the cause of the low accuracy of the national scale land-use scenarios as well as priority solutions, such as aligning the underlying spatial vegetation maps and improving the model to reduce two types of disagreement between the simulation and reference maps. These findings should help to improve the accuracy of model predictions and help to better inform policymakers during the decision-making process.

Assessing the Impact of Neighborhood Size on Temporal Convolutional Networks for Modeling Land Cover Change

“…Therefore, three-map comparison measures have been used to differentiate changed and persistent areas by comparing an initial map representing the real-world LC states at time t 0 , a reference map representing the real-world LC at time t n , and the forecasted LC map for time t n [60]. Measures that involve three-map comparison include Figure of Merit (FOM), Producer's Accuracy (PA), and User's Accuracy (UA), as expressed in previous work [62][63][64][65] and in Appendix A. FOM indicates the ratio of correctly forecasted changes versus the total amount of real-world and projected changes. Derived from components of FOM (Table A1), PA expresses the area forecasted correctly as changed versus the real-world changed areas, while UA expresses the amount of correctly changed area versus the real-world changed areas.…”

“…To investigate the types of spatial errors associated with each forecast, the total error or disagreement between real-world change and forecasted change can be explored with respect to error due to quantity (EQ) and error due to allocation (EA), as expressed in prior works [64][65][66] and in Appendix A. EQ indicates the difference in forecasted and real-world changed area, while EA indicates the amount of area allocated incorrectly as changed or to the wrong LC class. Both are expressed in terms of disagreeing area.…”

Scenario-Based Land Use and Land Cover Change Detection and Prediction Using the Cellular Automata–Markov Model in the Gumara Watershed, Upper Blue Nile Basin, Ethiopia

“…The prediction of future LULC dynamics based on historical trends is a critical issue [23]. However, the assumption of the continuity of the current land use trend is mostly susceptible to uncertainties due to the dynamic nature of land use change driver variables [24,25]. For this purpose, scenario-based LULC change analysis is mostly recommended as a tool to explore such uncertainties [26].…”

Quantitative Morphometric 3D Terrain Analysis of Japan Using Scripts of GMT and R

“…The performance of various GIS is, in many cases, tailored to their specific tasks, among which vector and raster spatial data processing is arguably the most prominent and important functionality of these software. For instance, some are better suited for image processing, such as Erdas Imagine [1], Idrisi GIS [2,3], Integrated Land and Water Information System (ILWIS) GIS [4], and ENvironment for Visualizing Images (ENVI) GIS, while others are best at vector data analysis and visualization, e.g., ArcGIS [5][6][7][8][9]. Most of these GIS are based on a standard interface with a restricted functionality that requires the manual processing of data, although recently machine learning techniques have been applied to spatial data modelling [10] and image analysis [11,12].…”