|Liya Fan, Betti Ketzer and Prof. Dr. Christian Bernhofer from the Department of Hydrology and Meteorology at the TU Dresden examined the spectral reflectivity of grassland within the spectral range of 380 nm up to 1700 nm in Inner Mongolia. |
They used two diode array spectrometers getSpec-PDA-UV/VIS and getSpec-NIR-1.7.
In the context of the research group 536 “Matter fluxes in grasslands of Inner Mongolia as influenced by stocking rate (MAGIM, sponsor number BE 172/7-1)” – an interdisciplinary German-Chinese cooperation project which is sponsored by the German Study Group (DFG) – two spectrometers of the company getSpec c/o Sentronic GmbH were used by the Department of Hydrology and Meteorology (IHM) at the TU Dresden to examine the influence of the grazing intensity on the spectral reflectivity of grassland within visible and near-infrared.
The examination area of the MAGIM project, the drainage basin of the Xilin River, is located in Inner Mongolia – a self-governed region in the Northeast of China. The exact examination spaces are situated close to the Inner Mongolian Grassland Ecosystem Research Station (IMGERS, 116░42' E, 43░38' N) which is directed by the Institute of Botany of the Chinese Academy of Science (IB-CAS).
The reflection measurements were made by using the following equipement:
- UV/VIS-Spectrometer getSpec-PDA, 380-11100 nm, FWHM 1.1nm
- NIR-Spectrometer getSpec-NIR-1.7-128-TS, 900-1700 nm, FWHM 12 nm
- Y-optical fiber with 400 Ám core diameter
- collimating lens Col-UV/VIS
- white standard Spectralon (cp. getReflex) with 99% reflexion as reference material
The measurement setup is described in figure 1. The sun light illuminated the meadow. The white reflection standard has been pivoted so that it was inserted into the optical path just for reflection measurement. A 12V/12A battery supplied power for both spectrometers.
Figure 1: Measurement setup for reflection measurement on a meadow in Inner Mongolia
Performance and Analysis:
The spectrometer measurements were performed from May to September on five spaces with a different grazing intensity (sheep and goats). Measurements were performed on two spaces which had not been grazed since 1979 and 1999 respectively and on one moderately grazed space (approx. 1.2 sheep units per hectare, a single sheep unit conforms to a dam with a pup). Furthermore measurements were performed on a spot used from mid of Oktober to mid of February (winter feedlot) as well as on an overgrazed space (approx. 2 sheep unit per hectare).
The analysis of the measured spectral reflections included the creation of time series for the spaces in detail (see figure 1 with example of the overgrazed space) as well as the comparison of the differencies between the examination spaces (see figure 2, comparison of the spaces for example on August 17th, 2005).
Figure 2: Time series of spectral reflection of the overgrazed space
Figure 3: Comparison of spectral reflections concerning five spaces with different grazing intensity
By means of the measurements of the spectral reflection of different spots it was possible to prove that the plant growth on the different spaces changes in the course of the year. Whereas at the beginning (May and June) as well as at the end of the vegetation period (September) higher spectral reflections were measured, the values are lower in the mid of the vegetation period (July and August). Furthermore differences between the single examination spaces were verified.
The spectrometer measurements are intended for the derivation of a vegetation index – the NDVI (Normalized Vegetation Index) – in the near future. This index is a measure for the photosynthetic activity of plants. If the vegetation density is constant for an area, the NDVI can also be used as an indicator for the condition and growth intensity of the vegetation. The NDVI can be easily derivated from remote sensing values as well. A comparison of the NDVI measured on the ground and accordant sattelite values is projected. Last but not least the spectral reflectance measured in situ are intendend to be used for the validation of remote sensing data.