Modeling multiple land use changes using ANN, CART and MARS: Comparing tradeoffs in goodness of fit and explanatory power of data mining tools

Tayyebi, Amin; Pijanowski, Bryan C.

doi:10.1016/j.jag.2013.11.008

Cited by 150 publications

(88 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The values p 1 and p 0 are the proportions of correctly classified test pixels of two classifiers under comparison. In addition, receiver operating characteristic (ROC) curves and the corresponding AUC have been calculated; the AUC is an increasingly used accuracy metric in machine-learning and data mining [102][103][104]. The AUC ranges from 0-100%, with 100% representing an error-free classification.…”

Section: Accuracy Assessmentmentioning

confidence: 99%

Mapping Cropland Abandonment in the Aral Sea Basin with MODIS Time Series

Löw¹,

Prishchepov

Waldner

et al. 2018

Remote Sensing

View full text Add to dashboard Cite

Abstract:Cropland abandonment is globally widespread and has strong repercussions for regional food security and the environment. Statistics suggest that one of the hotspots of abandoned cropland is located in the drylands of the Aral Sea Basin (ASB), which covers parts of post-Soviet Central Asia, Afghanistan and Iran. To date, the exact spatial and temporal extents of abandoned cropland remain unclear, which hampers land-use planning. Abandoned land is a potentially valuable resource for alternative land uses. Here, we mapped the abandoned cropland in the drylands of the ASB with a time series of the Normalized Difference Vegetation Index (NDVI) from the Moderate Resolution Imaging Spectroradiometer (MODIS) from [2003][2004][2005][2006][2007][2008][2009][2010][2011][2012][2013][2014][2015][2016]. To overcome the restricted ability of a single classifier to accurately map land-use classes across large areas and agro-environmental gradients, "stratum-specific" classifiers were calibrated and classification results were fused based on a locally weighted decision fusion approach. Next, the agro-ecological suitability of abandoned cropland areas was evaluated. The stratum-specific classification approach yielded an overall accuracy of 0.879, which was significantly more accurate (p < 0.05) than a "global" classification without stratification, which had an accuracy of 0.811. In 2016, the classification results showed that 13% (1.15 Mha) of the observed irrigated cropland in the ASB was idle (abandoned). Cropland abandonment occurred mostly in the Amudarya and Syrdarya downstream regions and was associated with degraded land and areas prone to water stress. Despite the almost twofold population growth and increasing food demand in the ASB area from 1990 to 2016, abandoned cropland was also located in areas with high suitability for farming. The map of abandoned cropland areas provides a novel basis for assessing the causes leading to abandoned cropland in the ASB. This contributes to assessing the suitability of abandoned cropland for food or bioenergy production, carbon storage, or assessing the environmental trade-offs and social constraints of recultivation.

show abstract

Section: Accuracy Assessmentmentioning

confidence: 99%

Mapping Cropland Abandonment in the Aral Sea Basin with MODIS Time Series

Löw¹,

Prishchepov

Waldner

et al. 2018

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…A decision tree algorithm is a top-down classification approach which is used to organize complex data sets step-by-step by effectively combining their characteristics with professional knowledge [45,50]. A decision tree comprises root, internal, and leaf nodes; of these, parent nodes (i.e., root or internal nodes) are further classified as child nodes (i.e., internal or leaf nodes) based on classification rules.…”

Section: Decision Tree Algorithmmentioning

confidence: 99%

Spatiotemporal Patterns of Crop Irrigation Water Requirements in the Heihe River Basin, China

Liu

Song

Deng

2017

Water

View full text Add to dashboard Cite

Abstract:Agricultural expansion, population growth, rapid urbanization, and climate change have all significantly impacted global water supply and demand and have led to a number of negative consequences including ecological degradation and decreases in biodiversity, especially in arid and semi-arid areas. The agricultural sector consumes the most water globally; crop irrigation alone uses up more than 80% of available agricultural water. Thus, to maintain sustainable development of the global economy and ecosystems, it is crucial to effectively manage crop irrigation water. We focus on the arid and semi-arid Heihe River Basin (HRB), China, as a case study in this paper, extracting spatiotemporal information on the distribution of crop planting using multi-temporal Thematic Mapper and Enhanced Thematic Mapper Plus (TM/ETM+) remote sensing (RS) images. We estimate the spatiotemporal crop irrigation water requirements (IWR c ) using the Food and Agriculture Organization of the United Nations (FAO) Penman-Monteith method and reveal variations in IWR c . We also analyze the impact of changes in crop planting structure on IWR c and discuss strategies for the rational allocation of irrigation water as well as policies to alleviate imbalance between water supply and demand. The results of this study show that effective rainfall (ER) decreases upstream-to-downstream within the HRB, while crop evapotranspiration under standard conditions (ET c ) increases, leading to increasing spatial variation in IWR c from zero up to 150 mm and between 300 and 450 mm. Data show that between 2007 and 2012, annual mean ER decreased from 139.49 to 106.29 mm, while annual mean ET c increased from 483.87 to 500.38 mm, and annual mean IWR c increased from 339.95 to 370.11 mm. Data show that monthly mean IWR c initially increased before decreasing in concert with crop growth. The largest values for this index were recorded during the month of June; results show that IWR c for May and June decreased by 8.14 and 11.67 mm, respectively, while values for July increased by 5.75 mm between 2007 and 2012. These variations have helped to ease the temporal imbalance between water supply and demand. Mean IWR c values for oilseed rape, corn, barley, and other crops all increased over the study period, from 208. 43, 349.35, 229.26, and 352.85 mm, respectively, in 2007, to 241.81, 393.10, 251.17, and 378.86 mm, respectively, in 2012. At the same time, the mean IWR c of wheat decreased from 281.53 mm in 2007 to 266.69 mm in 2012. Mainly because of changes in planting structure, the total IWR c for the HRB in 2012 reached 2692.58 × 10 6 m 3 , an increase of 332.16 × 10 6 m 3 (14.07%) compared to 2007. Data show that 23.11% (76.77 × 10 6 m 3 ) of this increase is due to crop transfers, while the remaining 76.89% (255.39 × 10 6 m 3 ) is the result of the rapid expansion of cultivated land. Thus, to maintain both the sustainable development and ecological security of the HRB, it is crucial to efficiently manage and utilize agricultural water in light of spat...

show abstract

“…Pasolli, Truong, Malik, Waldron, & Segata, 2016; Shipp et al., 2002; Tango & Botta, 2013), there has been a recent increase in its use within the geospatial and ecological sciences. For instance, ML has been successfully applied to predict species distribution (Liu, White, Newell, & Griffioen, 2013), land‐use change (Tayyebi & Pijanowski, 2014), and hydrological regimes (Cross et al., 2015) and has also been applied to vegetation mapping across a range of spatial scales using a variety of algorithms (e.g. Bradter, Thom, Altringham, Kunin, & Benton, 2011; Munyati, Ratshibvumo, & Ogola, 2013; Pesch, Schmidt, Schroeder, & Weustermann, 2011; Zhang & Xie, 2013).…”

Section: Introductionmentioning

confidence: 99%

Impact of ecological redundancy on the performance of machine learning classifiers in vegetation mapping

MacIntyre

Niekerk

Dobrowolski

et al. 2018

Ecology and Evolution

View full text Add to dashboard Cite

Vegetation maps are models of the real vegetation patterns and are considered important tools in conservation and management planning. Maps created through traditional methods can be expensive and time‐consuming, thus, new more efficient approaches are needed. The prediction of vegetation patterns using machine learning shows promise, but many factors may impact on its performance. One important factor is the nature of the vegetation–environment relationship assessed and ecological redundancy. We used two datasets with known ecological redundancy levels (strength of the vegetation–environment relationship) to evaluate the performance of four machine learning (ML) classifiers (classification trees, random forests, support vector machines, and nearest neighbor). These models used climatic and soil variables as environmental predictors with pretreatment of the datasets (principal component analysis and feature selection) and involved three spatial scales. We show that the ML classifiers produced more reliable results in regions where the vegetation–environment relationship is stronger as opposed to regions characterized by redundant vegetation patterns. The pretreatment of datasets and reduction in prediction scale had a substantial influence on the predictive performance of the classifiers. The use of ML classifiers to create potential vegetation maps shows promise as a more efficient way of vegetation modeling. The difference in performance between areas with poorly versus well‐structured vegetation–environment relationships shows that some level of understanding of the ecology of the target region is required prior to their application. Even in areas with poorly structured vegetation–environment relationships, it is possible to improve classifier performance by either pretreating the dataset or reducing the spatial scale of the predictions.

show abstract

Modeling multiple land use changes using ANN, CART and MARS: Comparing tradeoffs in goodness of fit and explanatory power of data mining tools

Cited by 150 publications

References 34 publications

Mapping Cropland Abandonment in the Aral Sea Basin with MODIS Time Series

Mapping Cropland Abandonment in the Aral Sea Basin with MODIS Time Series

Spatiotemporal Patterns of Crop Irrigation Water Requirements in the Heihe River Basin, China

Impact of ecological redundancy on the performance of machine learning classifiers in vegetation mapping

Contact Info

Product

Resources

About