Digitally Mapping Gypsic and Natric Soil Areas Using Landsat ETM Data

Abstract

Mapping salt‐affected soils in remote rangelands is challenging. We used Landsat 7 ETM data to facilitate digital mapping of gypsic and natric soil areas in the upper Colorado River drainage. Optimum index factor band combinations were used to explore the scene. Normalized difference ratio models and threshold values were developed by comparing spectral signatures with gypsic and natric soil areas verified in the field. Gypsic soil areas were mapped using the normalized difference ratio of Bands 5 and 7 with a threshold >0.11, probably related to the spectral reflectance of gypsum within a few centimeters of the surface. All sites predicted to be gypsic soil areas were determined to be gypsic by field assessment, and 87% of the field‐observed gypsic soil areas were correctly predicted. Natric soil areas were mapped using the normalized difference ratio of Bands 5 and 4 with a threshold >0.19, possibly related to the co‐occurrence of Fe‐bearing minerals with natric soil areas. Most of the sites predicted to be natric were determined in the field to be natric (82%), but only half of the field‐observed natric areas were correctly predicted, indicating that natric soils are harder to detect spectrally than gypsic soils. While the gypsic model may be transferred to other areas, particularly in the arid Colorado Plateau, transfer of natric models would be difficult. Normalized difference ratio models can be developed for other digital soil mapping areas where land surface features produce differences in Landsat spectral band reflectances.

Random Forests Applied as a Soil Spatial Predictive Model in Arid Utah

“…The Landsat scene was standardized using the cosine theta (COST) method without tau (Chavez, 1996;RSGIS, 2003: script no. 3;Nield et al, 2007). The values for the dark object subtraction were sampled from Fish Lake, Utah, (deep lake) and shadows cast by cumulus clouds.…”

“…The values for the dark object subtraction were sampled from Fish Lake, Utah, (deep lake) and shadows cast by cumulus clouds. These 5 and 1, 4 and 7, and 3 and 1 exhibited unique patterns wherein distinct landforms and vegetation communities were visually identified and thought to be useful in the model (Cole, 2004;Bodily, 2005;Scull et al, 2005;Nield et al, 2007;Saunders and Boettinger, 2007) (Figure 7, 8, and 9). …”

Modeling and Mapping of Soil Salinity with Reflectance Spectroscopy and Landsat Data Using Two Quantitative Methods (PLSR and MARS)

“…Several studies have shown the visible, near infrared, or short-wave infrared spectral bands from the optical sensors to be promising for the detection of surface soil salinity [11][12][13][14][15]. In addition, hyperspectral data have been successfully used in several studies on soil salinity, enabling quantitative assessment of salt-affected soils [16][17][18][19][20][21][22].…”

An heuristic uncertainty directed field sampling design for digital soil mapping

“…The key to selecting environmental covariates is that the covariate reflects the spatial variation of soils and at the same time data on spatial variation of this covariate can be easily obtained. With the development of earth observation technologies and spatial analysis method, many new useful covariates are emerging (Nield et al, 2007;Zhu et al (2010b); Liu et al, 2012;Qin et al, 2012). Statistical methods (such as principal components analysis, stepwise regression) could help us select appropriate covariates (Pechenizkiy et al, 2003;Behrens et al, 2010;Samuel-Rose et al, 2015;Ramadan et al, 2001;Mansuy et al, 2014;Beaudoin et al, 2014).…”