Terrace extraction based on remote sensing images and digital elevation model in the loess plateau, China

Luo, Lanhua; Li, Fayuan; Dai, Ziyang; Xue, Yan; Liu, Wei; Fang, Xuan

doi:10.1007/s12145-020-00444-x

Cited by 21 publications

(12 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The accuracy evaluation result using the 301 test samples of known terraces (Table 4) was numerically similar to the above result using the 10 875 random test samples (Table 3). The Chi-square tests (Mantel, 1963) were carried out for the two PAs and UAs of terrace class, respectively, to further prove the similarity quantitatively. The p values of both tests were greater than 0.05, indicating there was no statistically significant difference between the terrace accuracies using the two test sample sets.…”

Section: Accuracy Assessmentmentioning

confidence: 99%

A 30 m terrace mapping in China using Landsat 8 imagery and digital elevation model based on the Google Earth Engine

Cao

Naipal

et al. 2021

Earth Syst. Sci. Data

View full text Add to dashboard Cite

Abstract. The construction of terraces is a key soil conservation practice on agricultural land in China providing multiple valuable ecosystem services. Accurate spatial information on terraces is needed for both management and research. In this study, the first 30 m resolution terracing map of the entire territory of China is produced by a supervised pixel-based classification using multisource and multi-temporal data based on the Google Earth Engine (GEE) platform. We extracted time-series spectral features and topographic features from Landsat 8 images and the Shuttle Radar Topography Mission digital elevation model (SRTM DEM) data, classifying cropland area (cultivated land of Globeland30) into terraced and non-terraced types through a random forest classifier. The overall accuracy and kappa coefficient were evaluated by 10 875 test samples and achieved values of 94 % and 0.72, respectively. For terrace class, the producer's accuracy (PA) was 79.945 %, and the user's accuracy (UA) was 71.149 %. The classification performed best in the Loess Plateau and southwestern China, where terraces are most numerous. Some northeastern, eastern-central, and southern areas had relatively high uncertainty. Typical errors in the mapping results are from the sloping cropland (non-terrace cropland with a slope of ≥ 5∘), low-slope terraces, and non-crop vegetation. Terraces are widely distributed in China, and the total terraced area was estimated to be 53.55 Mha (i.e., 26.43 % of China's cropland area) by pixel counting (PC) method and 58.46 ± 2.99 Mha (i.e., 28.85 % ± 1.48 % of China's cropland area) by error-matrix-based model-assisted estimation (EM) method. Elevation and slope were identified as the main features in the terrace/non-terrace classification, and multi-temporal spectral features (such as percentiles of NDVI, TIRS2, and BSI) were also essential. Terraces are more challenging to identify than other land use types because of the intra-class feature heterogeneity, interclass feature similarity, and fragmented patches, which should be the focus of future research. Our terrace mapping algorithm can be used to map large-scale terraces in other regions globally, and our terrace map will serve as a landmark for studies on multiple ecosystem service assessments including erosion control, carbon sequestration, and biodiversity conservation. The China terrace map is available to the public at https://doi.org/10.5281/zenodo.3895585 (Cao et al., 2020).

show abstract

Section: Accuracy Assessmentmentioning

confidence: 99%

A 30 m terrace mapping in China using Landsat 8 imagery and digital elevation model based on the Google Earth Engine

Cao

Naipal

et al. 2021

Earth Syst. Sci. Data

View full text Add to dashboard Cite

show abstract

“…Since the end of 1990s, the emergence and development of digital terrain analysis based on digital elevation models (DEMs) provided important conditions for the research of geomorphology, especially on large scales (Wilson and Gallant 2000;Florinsky 2002;Li et al 2005b;Tang et al 2005;Florinsky 2011;Wilson 2012;Li et al 2016;Luo et al 2020). The key of digital terrain analysis is extraction and geostatistical analysis of the topographic feature elements (including topographic feature point, line, and surface) and topographic factors (such as slope gradient, slope direction, plane curvature, profile curvature, topographic relief, coefficient of elevation variation, surface cutting depth) based on DEMs (Oostwoud Wijdenes et al 2000;Hancock and Evans 2006;McNamara et al 2006;Yang et al 2009;Tang 2014;Torri et al 2018).…”

Section: Introductionmentioning

confidence: 99%

Spatial variation of gully development in the loess plateau of China based on the morphological perspective

Dai

et al. 2020

Earth Sci Inform

Self Cite

View full text Add to dashboard Cite

Gullies characterized by frequent material exchange have profoundly affected the loess landforms evolution in the Loess Plateau of China. Accompanying gully development, various types of gullies have shaped the distinctive loess landforms. Thus, the spatial variations of gully morphology map the regional differences of gully development and loess landforms. Based on 6 key test areas with 5 m resolution DEM and 156 evenly distributed test areas with 25 m resolution DEM, a comprehensive gullyshape index system of topographic feature points, lines, and surfaces was constructed to describe spatial differences of the gully development. After comparing gully-shape indexes in the key test areas, an obvious correlation existed between these indexes and loess landform types. The gully development degree was low in Shenmu and Chunhua, high in Suide, Yanchuan, and Ganquan, while Yijun lay in-between. Using the method of Universal Kriging interpolation, a series of spatial interpolation maps of the gully-shape indexes were obtained, which revealed gully development degree in the whole Loess Plateau. Overall, The active areas of gully development were basically distributed along loess hilly ridges, loess hills, loess ridges, and in a few medium mountain areas, while the inactive areas were basically distributed in the thin loess coverage areas, including plains, river terraces, aeolian dunes, and a few loess platforms. Finally, the gully development regionalization was realized, which contained 6 subzones. These results provided significant references for loess landform research and soil erosion management in the Loess Plateau.

show abstract

“…Several studies on land cover mapping have shown that adding textural variables can increase the mapping accuracy (Johansen et al, 2007;Masjedi et al, 2016;Rodriguez-Galiano et al, 2012). Moreover, some small-scale research has shown the effectiveness of textures in terrace/non-terrace classification (Li et al, 2013;Luo et al, 2020;Zhang et al, 2017). The ability of textures to discriminate different land cover types relates to the image spatial resolution (Chen et al, 2004).…”

Section: Limitations and Directions For Future Researchmentioning

confidence: 99%

“…As for the selected data for classification of terrace/non-terrace, usually, past studies relied on images from high-resolution satellites such as WorldView (Luo et al, 2020;Zhao et al, 2017), SPOT-5 (Li et al, 2013), Gaofen-1 (Zhang et al, 2017) and unmanned aerial vehicles (Diaz-Varela et al, 2014). Such data are expensive and are not universally accessible, making them unsuitable for large-scale applications.…”

Section: Introductionmentioning

confidence: 99%

A 30-meter terrace mapping in China using Landsat 8 imagery and digital elevation model based on the Google Earth Engine

Cao¹,

Yu²,

Naipal³

et al. 2020

Preprint

View full text Add to dashboard Cite

Abstract. The construction of terraces is a key soil conservation practice on agricultural land in China, providing multiple valuable ecosystem services. Accurate spatial information on terraces is needed for both management and research. In this study, the first 30 m resolution terracing map of the entire territory of China is produced by a supervised pixel-based classification using multi-source and multi-temporal data based on the Google Earth Engine (GEE) platform. We extracted time-series spectral features and topographic features from Landsat 8 images and the Shuttle Radar Topography Mission digital elevation model (SRTM DEM) data, classifying cropland area (cultivated land of Globeland30) into terraced and non-terraced type through a random forest classifier. The overall accuracy and kappa coefficient were evaluated by 10875 test samples and achieved values of 94 % and 0.72, respectively. The classification performed best in the Loess Plateau and southwestern China, where terraces are most numerous. Some northeastern, central eastern and southern area had relatively high uncertainty. Typical errors in the mapping results from the sloping cropland (non-terrace cropland with a slope of ≥ 5°), low-slope terraces, and non-crop vegetation. Terraces are widely distributed in China and the total terraced area was estimated to be 53.55 Mha (i.e., 26.43 % of China's cropland area) by pixel counting (PC) method and 58.46 ± 2.99 Mha (i.e., 28.85 % ± 1.48 % of China's cropland area) by error matrix-based model-assisted estimation (EM) method. Elevation and slope were identified as the main features in the terrace/non-terrace classification, and multi-temporal spectral features (such as percentiles of NDVI, TIRS2, BSI) were also essential. Terraces are more challenging to identify than other land use types because of the intra-class feature heterogeneity, inter-class feature similarity and fragmented patches, which should be the focus of future research. Our terrace mapping algorithm can be used to map large-scale terraces in other regions globally, and our terrace map will serve as a landmark for studies on multiple ecosystem services assessments including erosion control, carbon sequestration, and biodiversity conservation. The China terrace map is available to the public at https://doi.org/10.5281/zenodo.3895585 (Cao et al., 2020).

show abstract

Terrace extraction based on remote sensing images and digital elevation model in the loess plateau, China

Cited by 21 publications

References 26 publications

A 30 m terrace mapping in China using Landsat 8 imagery and digital elevation model based on the Google Earth Engine

A 30 m terrace mapping in China using Landsat 8 imagery and digital elevation model based on the Google Earth Engine

Spatial variation of gully development in the loess plateau of China based on the morphological perspective

A 30-meter terrace mapping in China using Landsat 8 imagery and digital elevation model based on the Google Earth Engine

Contact Info

Product

Resources

About