Hyper-temporal remote sensing for digital soil mapping: Characterizing soil-vegetation response to climatic variability

Maynard, Jonathan J.; Levi, Matthew R.

doi:10.1016/j.geoderma.2016.09.024

Cited by 78 publications

(52 citation statements)

References 86 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Using comprehensive environmental geodata for DSM improves prediction accuracy because soil-forming factors are likely better represented by a larger number of covariates. Derivatives of geological or legacy soil maps (Nussbaum et al, 2014), multi-scale terrain analysis (Behrens et al, 2010a(Behrens et al, , b, 2014Miller et al, 2015), wide ranges of climatic parameters (Liddicoat et al, 2015) and (multi-temporal) imaging spectroscopy (Mulder et al, 2011;Poggio et al, 2013;Viscarra Rossel et al, 2015;Fitzpatrick et al, 2016;Hengl et al, 2017;Maynard and Levi, 2017) all contribute to generating high-dimensional sets of partly multi-collinear covariates. One usually presumes that DSM techniques benefit from a large number of covariates even if a method selects only a small subset of relevant covariates for creating the predictions.…”

Section: Introductionmentioning

confidence: 99%

Evaluation of digital soil mapping approaches with large sets of environmental covariates

Nussbaum¹,

Spiess²,

Baltensweiler³

et al. 2018

SOIL

205

View full text Add to dashboard Cite

Abstract. The spatial assessment of soil functions requires maps of basic soil properties. Unfortunately, these are either missing for many regions or are not available at the desired spatial resolution or down to the required soil depth. The field-based generation of large soil datasets and conventional soil maps remains costly. Meanwhile, legacy soil data and comprehensive sets of spatial environmental data are available for many regions.Digital soil mapping (DSM) approaches relating soil data (responses) to environmental data (covariates) face the challenge of building statistical models from large sets of covariates originating, for example, from airborne imaging spectroscopy or multi-scale terrain analysis. We evaluated six approaches for DSM in three study regions in Switzerland (Berne, Greifensee, ZH forest) by mapping the effective soil depth available to plants (SD), pH, soil organic matter (SOM), effective cation exchange capacity (ECEC), clay, silt, gravel content and fine fraction bulk density for four soil depths (totalling 48 responses). Models were built from 300-500 environmental covariates by selecting linear models through (1) grouped lasso and (2) an ad hoc stepwise procedure for robust external-drift kriging (georob). For (3) geoadditive models we selected penalized smoothing spline terms by component-wise gradient boosting (geoGAM). We further used two tree-based methods: (4) boosted regression trees (BRTs) and (5) random forest (RF). Lastly, we computed (6) weighted model averages (MAs) from the predictions obtained from methods 1-5.Lasso, georob and geoGAM successfully selected strongly reduced sets of covariates (subsets of 3-6 % of all covariates). Differences in predictive performance, tested on independent validation data, were mostly small and did not reveal a single best method for 48 responses. Nevertheless, RF was often the best among methods 1-5 (28 of 48 responses), but was outcompeted by MA for 14 of these 28 responses. RF tended to over-fit the data. The performance of BRT was slightly worse than RF. GeoGAM performed poorly on some responses and was the best only for 7 of 48 responses. The prediction accuracy of lasso was intermediate. All models generally had small bias. Only the computationally very efficient lasso had slightly larger bias because it tended to under-fit the data. Summarizing, although differences were small, the frequencies of the best and worst performance clearly favoured RF if a single method is applied and MA if multiple prediction models can be developed.Published by Copernicus Publications on behalf of the European Geosciences Union.

show abstract

Section: Introductionmentioning

confidence: 99%

Evaluation of digital soil mapping approaches with large sets of environmental covariates

Nussbaum¹,

Spiess²,

Baltensweiler³

et al. 2018

SOIL

205

View full text Add to dashboard Cite

show abstract

“…RS has greatly contributed to SOC mapping (Grinand et al, 2017;Maynard & Levi, 2017). To integrate multivariate factors acquired using RS data, random forest (RF) models can be employed.…”

Section: Introductionmentioning

confidence: 99%

Large‐scale soil organic carbon mapping based on multivariate modelling: The case of grasslands on the Loess Plateau

Wang

Deng

et al. 2017

Land Degrad Dev

View full text Add to dashboard Cite

The Loess Plateau is considered one of the world's regions with severe soil erosion. Grasslands are widely distributed on the Loess Plateau, accounting for approximately 40% of the total area.Soil organic carbon (SOC) plays an important role in the terrestrial carbon cycle in this region.We compiled more than 1,000 measurements of plant biomass and SOC content derived from 223 field studies of grasslands on the Loess Plateau. Combined with meteorological factors (precipitation and air temperature) and the photosynthetically active radiation factor, the topsoil SOC contents of grasslands were predicted using the random forest (RF) regression algorithm. biomass and temperature should receive more attention due to increasing C sequestration.

show abstract

“…[17]]. However, the single image approach has been criticized for lacking information on intra-seasonal growth dynamics [24]. On the other hand, multitemporal images and the derived phenological metrics uncover the intra-annual biophysical properties of the crop across the field, as driven by soil-climate interaction.…”

Section: Summary Of Indicative Phenological Metrics For Crop Field Mamentioning

confidence: 99%

“…In precision agriculture, NDVI has been used as a surrogate for crop yield for SSCM zone delineation [17] [22] [23]. While single images are useful for yield estimation, the inter annual growth variability resulting from soil-climate interaction can produce spurious results depending on the image date selection [24]. Multi-temporal NDVI, on the other hand, provides an additional temporal dimension to uncover the vegetation dynamics.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Remote Sensing Derived Phenological Metrics to Assess the Spatio-Temporal Growth Variability in Cropping Fields

Araya¹,

Ostendorf²,

Lyle³

et al. 2017

ARS

View full text Add to dashboard Cite

Precision Agriculture (PA) recognizes and manages intra-field spatial variability to increase profitability and reduced environmental impact. Site Specific Crop Management (SSCM), a form of PA, subdivides a cropping field into uniformly manageable zones, based on quantitative measurement of yield limiting factors. In Mediterranean environments, the spatial and temporal yield variability of rain-fed cropping system is strongly influenced by the spatial variability of Plant Available Water-holding Capacity (PAWC) and its strong interaction with temporally variable seasonal rainfall. The successful adoption of SSCM depends on the understanding of both spatial and temporal variabilities in cropping fields. Remote sensing phenological metrics provide information about the biophysical growth conditions of crops across fields. In this paper, we examine the potential of phenological metrics to assess the spatial and temporal crop growth variability across a cropping field. The study was conducted at a wheat cropping field in Minnipa, South Australia. The field was classified into three management zones using prolonged observations including soil assessment and multiple year yield data. The main analytical steps followed in this study were: calculation of the phenological metrics using time series NDVI data from Moderate Resolution Imaging Spectroradiometer (MODIS) for 15 years (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015); producing spatial trend and temporal variability maps of phenological metrics; and finally, assessment of association between the spatial patterns and temporal variability of the metrics with management zones of the cropping field. The spatial trend of the seasonal peak NDVI metric showed significant association with the management zone pattern. In terms of temporal variability, Time-integrated NDVI (TINDVI) showed higher variability in the "good" zone compared with the "poor" zone. This indicates that the magnitude of the seasonal peak is more sensitive to soil related factors across the field, whereas TINDVI is more sensitive to seasonal variability. The interpretation of the association between phenological metrics and the management zone site conditions was discussed in relation to soil-climate interaction. The results demonstrate the potential of the phenological metrics to assess the spatial and temporal growth variability across cropping fields and to understand the soil-climate interaction. The approach presented in this paper provides a pathway to utilize phenological metrics for precision agricultural management application.

show abstract

Hyper-temporal remote sensing for digital soil mapping: Characterizing soil-vegetation response to climatic variability

Cited by 78 publications

References 86 publications

Evaluation of digital soil mapping approaches with large sets of environmental covariates

Evaluation of digital soil mapping approaches with large sets of environmental covariates

Large‐scale soil organic carbon mapping based on multivariate modelling: The case of grasslands on the Loess Plateau

Remote Sensing Derived Phenological Metrics to Assess the Spatio-Temporal Growth Variability in Cropping Fields

Contact Info

Product

Resources

About