A spatial system dynamic model for regional desertification simulation – A case study of Ordos, China
Introduction
Desertification is defined as land degradation in arid, semi-arid, and dry sub-humid areas resulting from various factors including climatic variations and human activities (UNCCD, 1994). The interaction of driving forces at different spatio-temporal scales leads to a highly complex desertification process (Prince, 2002, Peters and Havstad, 2006). Rapid urbanization, intensive land use, large-scale ecological projects, and significant climate variation have made desertification dynamics more complicated, especially in the past 30 years (Zhang et al., 2012, D'Odorico et al., 2013, Wang, 2014, Xu et al., 2014). Thus, modeling desertification dynamics as a function of climatic variations and human activities is necessary in order to predict desertification risk, evaluate the impact of climate change, and support policy-making for desertification rehabilitation.
Numerous studies have attempted to predict and simulate desertification dynamics at different temporal-spatial scales (Van Delden et al., 2007, Ibanes et al., 2008, Helldén, 2008, Feng et al., 2010, Rasmy et al., 2010). For example, some studies used regression models, Markov models, or Cellular Automata (CA) models to simulate desertification by considering the statistical relationship between desertification dynamics and driving forces in a certain historical period (Liu et al., 2008, Ding et al., 2009, Feng et al., 2010). Although these models have a robust statistical basis, they are limited in their ability to explain the process of desertification, especially when modeling the impact of social driving factors and supporting policy-making. Furthermore, these models assumed that these relationships will not change in the future. Desertified lands represent open and complex systems. Multi-level feedbacks and non-linear impacts are two important characteristics of the desertification driving process (Peters and Havstad, 2006, Veron et al., 2006). For example, the interactions in the livestock-grass-soil system are extremely important for desertification at small scales (Hahn et al., 2005, Ibanez et al., 2007), but the effect of these interactions will be enhanced or weakened within a large-scale analysis that incorporates climate variation and other human activities. With the development of systems theory, an increasing number of studies are attempting to simulate desertification using the method of System Dynamics (SD). As a classical SD model, the predator-prey model has been frequently used to construct the relationship between humans (predator) and natural resources (prey) in rangeland in order to stimulate desertification dynamics. In 1995, Puigdefábregas developed a desertification model using the predator-prey approach that considered the effect of a closed grazing system and human migration (Puigdefábregas, 1995). Due to its good scalability, the predator-prey model was gradually enlarged, especially the predator component (Puigdefábregas, 1998, Ibanes et al., 2008; Helldén, 2008; Rasmy et al., 2010). For example, Helldén (2008) developed a human-environment two-level coupled predator-prey model to simulate desertification dynamics in Sahelian; in this model, population stock was described as a function of growth rate, death rate, and resource-dependent migration. Rasmy et al. (2010) developed a dynamic simulation model of desertification in Egypt by considering the relationship among people, livestock, and plant cover. With the increasing demand for policy-support in regional and catchment environment management, simulation of desertification has often been included as one of the functions of a more comprehensive system models. Van Delden et al., 2007, Van Delden et al., 2009, Van Delden et al., 2011) designed a policy-support system for river basin management based on a multi-scale dynamic spatial model of socio-economic and physical processes, and this model included the simulation of land degradation and desertification.
Nevertheless, the SD method still faces some limitations when applied to desertification simulation. For example, several previous studies considered the research region as a whole when developing SD models for desertification simulation (Helldén, 2008, Rasmy et al., 2010). These models neglected the spatial heterogeneity of desertified lands as well as regional-scale driving forces. Furthermore, the degree of desertification was not considered in most simulations. Determining the desertification degree would aid in identifying the structural changes of desertified lands, assessing the effect of policies, and producing more detailed spatial planning for desertification control.
In contrast to classic system dynamic models, a spatial system dynamic (SSD) model can construct the linkage of SD variables to Geographical Information System (GIS), including both data association as well as semantic association (Zhang, 2008, Neuwirth et al., 2015). However, few studies have focused on developing an SSD model for desertification simulation at a regional scale that considers the effect of different natural and social factors (Van Delden et al., 2007). Therefore, this study attempts to develop an SSD model for desertification simulation by combining multi-level interactions between natural factors and socio-economic driving forces. We then apply the model to a typical region suffering desertification in China, the city of Ordos, to simulate desertification dynamics from 2011 to 2030. This model will hopefully support the desertification control policies of the local government.
Section snippets
Study area
The city of Ordos is located in the southwest of the Inner Mongolia Autonomous Region of China, with the range of 37° 41′–40° 51′ N and 106° 42′–111° 31′ E. Ordos has an area of approximately 86,752 km2 and had a population of roughly 1,480,000 in 2010 (Fig. 1). Ordos belongs to the arid and semi-arid transition zone, and is subject to a temperate continental climate. The annual rainfall ranges between 170 and 450 mm. The rainfall increases along a gradient from west to east, and the majority
Model framework
In contrast with other ecological and geographical parameters, desertification is only a description of land status, and it cannot be directly measured using conventional tools and units. Thus, it is necessary to select a common indicator that can reflect desertification status and that can be used as an intermediate variable to model the relationship between driving forces and desertification. Additionally, this indicator should be modeled using spatial data. As one of the most important
Model accuracy assessment
According to Table 5, the simulation accuracy of intermediate parameters is very high, especially for the total population. However, the simulation accuracy for total number of sheep and large animals is relatively lower. For the NPP simulation (Fig. 4), the R2 reached a maximum of 0.93, but the RMSE was relatively higher at approximately 28 gC/m2/a. As an indicator of desertification degree, the higher RMSE of NPP impacts the overall accuracy of desertification simulation. However, this RMSE
Discussion
As one of the most important land surface ecosystems, desertified land and its dynamics are very complex due to the interaction of various factors at different temporal and spatial scales. In particular, the high spatial heterogeneity of human activities that can exhibit either positive or negative impacts creates great uncertainty in the reversal and expansion processes of desertification (Prince, 2002, Evans and Geerken, 2004). From this perspective, the SSD model can fulfill the requirements
Conclusion
Desertification simulation is very important. It can be used to predict desertification dynamics in the future, and it can also be employed by policy-makers to evaluate and implement desertification control policies and projects. In this study, an SSD model was developed by incorporating spatial data into a traditional SD model. NPP was selected as a common indicator to measure the impacts and feedbacks of six driving forces and measure the different desertification grades. The application and
Acknowledgements
This research was jointly supported by the National Natural Science Foundation of China (Project 71573245 and 71103170), and the National Key Technologies R&D Program of China (2012BAC19B09).
References (48)
- et al.
Global desertification: drivers and feedbacks
Adv. Water Resour.
(2013) - et al.
Discrimination between climate and human-induced dryland degradation
J. Arid. Environ.
(2004) - et al.
Global net primary production: combining ecology and remote sensing
Remote Sens. Environ.
(1995) - et al.
A simulation model of long-term climate, livestock and vegetation interactions on communal rangelands in the semi-arid Succulent Karoo, Namaqualand, South Africa
Ecol. Model.
(2005) A coupled human-environment model for desertification simulation and impact studies
Glob. Planet. Change
(2008)- et al.
Effects of grassland restoration programs on ecosystems in arid and semiarid China
J. Environ. Manag.
(2013) - et al.
Desertification due to overgrazing in a dynamic commercial livestock-grass-soil system
Ecol. Model.
(2007) - et al.
Environmental decision support systems: current issues, methods and tools
Environ. Model. Softw.
(2007) - et al.
Assessing the impact of restoration-induced land conversion and management alternatives on net primary productivity in Inner Mongolian grassland, China
Glob. Planet. Change
(2013) - et al.
Modeling structural change in spatial system dynamics: a Daisyworld example
Environ. Model. Softw.
(2015)
Nonlinear dynamics in arid and semi-arid systems: interactions among drivers and processes across scales
J. Arid. Environ.
A dynamic simulation model of desertification in Egypt
Egypt. J. Remote Sens. Space Sci.
Stakeholder participation for environment management: a literature review
Biol. Conserv.
Remote sensing of crop production in China by production efficiency models: models comparisons, estimate and uncertainties
Ecol. Model.
Integration of multi-scale dynamic spatial models of socio-economic and physical processes for river basin management
Environ. Model. Softw.
A methodology for the design and development of integrated models for policy support
Environ. Model. Softw.
Assessing desertification
J. Arid. Environ.
Combating desertification in China: past, present and future
Land Use Policy
Multi-scale quantitative assessment of the relative roles of climate change and human activity in desertification-A case study of the Ordos Plateau, China
J. Arid. Environ.
The dynamics of desertification in the farming-pastoral region of North China over the past 10 years and their relationship to climate change and human activity
CATENA
Effectiveness of ecological restoration projects in Horqin Sandy Land, China based on SPOT-VGT NDVI data
Ecol. Eng.
Possible urban automata
Environ. Plan. B
Modeling urban dynamics through GIS-based cellular automata
Comput. Environ. Urban Syst.
A model for desertification evolution employing GIS with Cellular Automata
Int. Conf. Comput. Model. Simul. IEEE
Cited by (25)
The use of remote sensing for desertification studies: A review
2022, Journal of Arid EnvironmentsCitation Excerpt :This last definition has been formally and widely used, since then, for multiple studies on desertification around the world, thus providing multiple and variable scopes on how to measure, analyse, and model measure desertification. ( Kassas, 1995; Li et al., 2006; Cui et al., 2011; Bakr et al., 2012; Lamchin et al., 2016; Liu et al., 2018; Xu et al., 2016; Becerril-Piña et al., 2016; Zhao et al., 2018). Taking into consideration these definitions, for this study, we adopt the definition of desertification as: “land degradation in arid, semi-arid, and dry sub-humid areas resulting from human activities and climate variation which can lead to desert-like conditions”.
Path analysis model to identify and analyse the causes of aeolian desertification in Mu Us Sandy Land, China
2021, Ecological IndicatorsCitation Excerpt :Compared with long-time series monitoring in global large scale, regional studies could reveal driving mechanism more deeply (Jiang et al., 2019). The researches on driving mechanisms of aeolian desertification have been a hot topic for scholars (Xu et al., 2016). For instance, a previous study found that the aeolian desertification of El-Dakhla oasis in Egypt was mainly affected by climatic factors (Ismael, 2015).
Teamwork-oriented integrated modeling method for geo-problem solving
2019, Environmental Modelling and SoftwareCitation Excerpt :The earth's environmental system is a complex mix of natural and synthetic systems that includes massive and dynamic chemical, biological, and physical processes as well as complicated human activities (Gould et al., 2008; Walker and Chapra, 2014; Kay et al., 2015; Tao et al., 2019). Due to this complexity, various geo-problems – defined as problems that occur in geographic world – need to be explored, for example, environmental issues such as global warming, air pollution, and desertification (Bopp et al., 2005; Kiesewetter et al., 2015; Xu et al., 2016); decision-making and planning issues such as shortest path finding, shopping center site selection, and urban expansion planning (Önüt et al., 2010; Shahzada and Askar, 2011; Arsanjani et al., 2013); and human activity issues such as traffic congestion, information propagation via social media, and indoor evacuation (Sheu et al., 2004; He et al., 2015; Tang et al., 2016). Targeting the solution of different geo-problems, a range of geo-analysis models have been studied and developed in different domains (Al-Sabhan et al., 2003; Selvam et al., 2014; Yanosky et al., 2014; Basra et al., 2017; Shahparakia et al., 2017).
Exploring an efficient sandy barren index for rapid mapping of sandy barren land from Landsat TM/OLI images
2019, International Journal of Applied Earth Observation and GeoinformationCitation Excerpt :A decision tree approach has been applied to a combination of NDVI, topsoil grain size index and land surface albedo data to assess land cover change and desertification (Munkhnasan et al., 2016). Xu et al. (2016) developed a spatial system dynamic model for simulating desertification by integrating climate, soil water, population, economy, pasturage and land use data. The bare indexes and sandy barren detection methods described above were developed based on digital image analysis using the outputs of specific remote sensors and were not suitable for multi-source image inputs (inc. multi-spectral and hyper-spectral images).
Markov chains and cellular automata to predict environments subject to desertification
2018, Journal of Environmental Management