From local spectral measurements to maps of vegetation cover and biomass on the Qinghai-Tibet-Plateau: Do we need hyperspectral information?

Abstract Though the relevance of pasture degradation on the Qinghai-Tibet Plateau (QTP) is widely postulated, its extent is still unknown. Due to the enormous spatial extent, remote sensing provides the only possibility to investigate pasture degradation via frequently used proxies such as vegetation cover and aboveground biomass (AGB). However, unified remote sensing approaches are still lacking. This study tests the applicability of hyper- and multispectral in situ measurements to map vegetation cover and AGB on regional scales. Using machine learning techniques, it is tested whether the full hyperspectral information is needed or if multispectral information is sufficient to accurately estimate pasture degradation proxies. To regionalize pasture degradation proxies, the transferability of the locally derived ML-models to high resolution multispectral satellite data is assessed. 1183 hyperspectral measurements and vegetation records were performed at 18 locations on the QTP. Random Forests models with recursive feature selection were trained to estimate vegetation cover and AGB using narrow-band indices (NBI) as predictors. Separate models were calculated using NBI from hyperspectral data as well as from the same data resampled to WorldView-2, QuickBird and RapidEye channels. The hyperspectral results were compared to the multispectral results. Finally, the models were applied to satellite data to map vegetation cover and AGB on a regional scale. Vegetation cover was accurately predicted by Random Forest if hyperspectral measurements were used (cross validated R 2  = 0.89). In contrast, errors in AGB estimations were considerably higher (cross validated R 2  = 0.32). Only small differences in accuracy were observed between the models based on hyperspectral compared to multispectral data. The application of the models to satellite images generally resulted in an increase of the estimation error. Though this reflects the challenge of applying in situ measurements to satellite data, the results still show a high potential to map pasture degradation proxies on the QTP. Thus, this study presents robust methodology to remotely detect and monitor pasture degradation at high spatial resolutions.

[1]  Zhao Xinquan,et al.  GEO-ECOLOGICAL TRANSECT STUDIES IN NORTHEAST TIBET (QINGHAI, CHINA) REVEAL HUMAN-MADE MID-HOLOCENE ENVIRONMENTAL CHANGES IN THE UPPER YELLOW RIVER CATCHMENT CHANGING FOREST TO GRASSLAND , 2008 .

[2]  R Core Team,et al.  R: A language and environment for statistical computing. , 2014 .

[3]  E. Vermote,et al.  Validation of a vector version of the 6S radiative transfer code for atmospheric correction of satellite data. Part I: path radiance. , 2006, Applied optics.

[4]  Yong Zha,et al.  Assessment of grassland degradation near Lake Qinghai, West China, using Landsat TM and in situ reflectance spectra data , 2004 .

[5]  Thomas Foken,et al.  Pasture degradation modifies the water and carbon cycles of the Tibetan highlands , 2014 .

[6]  Max Kuhn,et al.  caret: Classification and Regression Training , 2015 .

[7]  B. Turner,et al.  Estimating foliage nitrogen concentration from HYMAP data using continuum, removal analysis , 2004 .

[8]  Shenggong Li,et al.  Grassland desertification by grazing and the resulting micrometeorological changes in Inner Mongolia , 2000 .

[9]  Jingyun Fang,et al.  Aboveground biomass in Tibetan grasslands , 2009 .

[10]  R. Long,et al.  Feed value of native forages of the Tibetan Plateau of China , 1999 .

[11]  Lukas W. Lehnert,et al.  A hyperspectral indicator system for rangeland degradation on the Tibetan Plateau: A case study towards spaceborne monitoring , 2014 .

[12]  K. Itten,et al.  Hyperspectral remote sensing for estimating aboveground biomass and for exploring species richness patterns of grassland habitats , 2011 .

[13]  Lukas W. Lehnert,et al.  Retrieval of grassland plant coverage on the Tibetan Plateau based on a multi-scale, multi-sensor and multi-method approach. , 2015 .

[14]  A. Skidmore,et al.  Red edge shift and biochemical content in grass canopies , 2007 .

[15]  Christoph Reudenbach,et al.  How old is pastoralism in Tibet? An ecological approach to the making of a Tibetan landscape , 2009 .

[16]  Luis Alonso,et al.  Machine learning regression algorithms for biophysical parameter retrieval: Opportunities for Sentinel-2 and -3 , 2012 .

[17]  Yang Yongping,et al.  Alpine steppe plant communities of the Tibetan highlands , 2011 .

[18]  Lukas W. Lehnert,et al.  Assessing pasture quality and degradation status using hyperspectral imaging: a case study from western Tibet , 2013, Remote Sensing.

[19]  Stefano Bocchi,et al.  Fine-scale assessment of hay meadow productivity and plant diversity in the European Alps using field spectrometric data , 2010 .

[20]  A. Skidmore,et al.  Hyperspectral band depth analysis for a better estimation of grass biomass (Cenchrus ciliaris) measured under controlled laboratory conditions , 2004 .

[21]  Prasad S. Thenkabail,et al.  Evaluation of Narrowband and Broadband Vegetation Indices for Determining Optimal Hyperspectral Wavebands for Agricultural Crop Characterization , 2002 .

[22]  Troy Sternberg,et al.  Piospheres and Pastoralists: Vegetation and Degradation in Steppe Grasslands , 2012 .

[23]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[24]  Daniela M. Witten,et al.  An Introduction to Statistical Learning: with Applications in R , 2013 .

[25]  L. Lehnert,et al.  Climate variability rather than overstocking causes recent large scale cover changes of Tibetan pastures , 2016, Scientific Reports.

[26]  Clement Atzberger,et al.  Estimation of Leaf Area Index Using DEIMOS-1 Data: Application and Transferability of a Semi-Empirical Relationship between two Agricultural Areas , 2013, Remote. Sens..

[27]  Shiro Itano,et al.  Reflectance spectra for monitoring green herbage mass in Zoysia‐dominated pastures , 2011 .

[28]  Jin Chen,et al.  Estimation of aboveground biomass using in situ hyperspectral measurements in five major grassland ecosystems on the Tibetan Plateau , 2008 .

[29]  Jay Gao,et al.  A spectral reflectance-based approach to quantification of grassland cover from Landsat TM imagery , 2003 .

[30]  Yaoming Ma,et al.  Assessment of vegetation dynamics and their response to variations in precipitation and temperature in the Tibetan Plateau , 2010 .

[31]  R. Harris.  Rangeland degradation on the Qinghai-Tibetan plateau: A review of the evidence of its magnitude and causes , 2010 .

[32]  Jian Sun,et al.  On the Variation of NDVI with the Principal Climatic Elements in the Tibetan Plateau , 2013, Remote. Sens..

[33]  Yue Shi,et al.  Phenology shift from 1989 to 2008 on the Tibetan Plateau: an analysis with a process-based soil physical model and remote sensing data , 2013, Climatic Change.

[34]  Qingzhu Gao,et al.  Alpine grassland degradation index and its response to recent climate variability in Northern Tibet, China , 2010 .

[35]  Douglas A. Johnson,et al.  Transformation of traditional pastoral livestock systems on the Tibetan steppe , 2006 .

[36]  Andy Liaw,et al.  Classification and Regression by randomForest , 2007 .

[37]  J. V. Soares,et al.  Evaluation of hyperspectral data for pasture estimate in the Brazilian Amazon using field and imaging spectrometers , 2008 .

[38]  Xinquan Zhao,et al.  Status and Dynamics of the Kobresia pygmaea Ecosystem on the Tibetan Plateau , 2008, Ambio.

[39]  Lukas W. Lehnert,et al.  Land Cover Change in the Andes of Southern Ecuador - Patterns and Drivers , 2015, Remote. Sens..

[40]  Max Kuhn,et al.  Applied Predictive Modeling , 2013 .

[41]  Roberta E. Martin,et al.  GRAZING SYSTEMS, ECOSYSTEM RESPONSES, AND GLOBAL CHANGE , 2004 .

[42]  S. Sandmeier,et al.  Physical Mechanisms in Hyperspectral BRDF Data of Grass and Watercress , 1998 .

[43]  Jan Hanspach,et al.  Plant communities of central Tibetan pastures in the Alpine Steppe / Kobresia pygmaea ecotone , 2011 .

[44]  Stephen D. Prince,et al.  Mapping land degradation by comparison of vegetation production to spatially derived estimates of potential production , 2008 .

[45]  R. Clark,et al.  Spectroscopic Determination of Leaf Biochemistry Using Band-Depth Analysis of Absorption Features and Stepwise Multiple Linear Regression , 1999 .

[46]  Nicholas M. Holden,et al.  The effects of enclosures and land-use contracts on rangeland degradation on the Qinghai–Tibetan plateau , 2013 .

[47]  R. Phillips,et al.  Estimating forage quantity and quality using aerial hyperspectral imagery for northern mixed-grass prairie , 2007 .

[48]  Berit Gehrke,et al.  Making Carex monophyletic (Cyperaceae, tribe Cariceae): a new broader circumscription , 2015 .