Several studies have demonstrated that soil reflectance decreases with increasing soil moisture content, or increases when the soil moisture reaches a certain content; however, there are few analyses on the quantitati...Several studies have demonstrated that soil reflectance decreases with increasing soil moisture content, or increases when the soil moisture reaches a certain content; however, there are few analyses on the quantitative relationship between soil reflectance and its moisture, especially in the case of black soils in northeast China. A new moisture adjusting method was developed to obtain soil reflectance with a smaller moisture interval to describe the quantitative relationship between soil reflectance and moisture. For the soil samples with moisture contents ranging from air-dry to saturated, the changes in soil reflectance with soil moisture can be depicted using a cubic equation. Both moisture threshold (MT) and moisture inflexion (MI) of soil reflectance can also be determined by the equation. When the moisture range was smaller than MT, soil reflectance can be simulated with a linear model. However, for samples with different soil organic matter (OM), the parameters of the linear model varied regularly with the OM content. Based on their relationship, the soil moisture can be estimated from soil reflectance in the black soil region.展开更多
Fraction of absorbed photosynthetically active radiation （FPAR） is one of the important variables in many vegetation productivity and biomass estimation models. Therefore, it is significant to retrieve FPAR accurate...Fraction of absorbed photosynthetically active radiation （FPAR） is one of the important variables in many vegetation productivity and biomass estimation models. Therefore, it is significant to retrieve FPAR accurately for the improvement of model precision. On the basis of the field experiment, this article analyzed the correlations between corn canopy FPAR and spectral reflectance, and reflectance derivative. Discussion about the mechanism of FPAR estimation with different empirical models is based upon corn canopy reflectance, reflectance derivative, NDVI （normalized difference vegetation index） and RVI （ratio vegetation index）. The reflectance of visible bands showed much better correlations with FPAR than near-infrared bands. The correlation between FPAR and reflectance derivative varied more frequently and greatly than that between FPAR and reflectance, and with preferable correlation only around 520, 570, 670, 805, 950, and 1 010 nm. Reflectance and reflectance derivative both had intimate correlation with FPAR at some typical single band, with the maximum R^2 of 0.791 and 0.882, respectively. In a word, reflectance derivative and vegetation index were much effective in the estimation of corn FPAR than reflectance, and the stepwise regression of multibands with reflectance derivative showed the best regression with R^2 of 0.944. Reflectance at 375 and 950 nm with absorption characteristics caused by water showed prodigious potential for FPAR precisely estimating model establishment. On the whole, vegetation index and reflectance derivative had good relationships with FPAR, and could be used for FAPR estimation. It would be effective for choosing right bands and excavating the hyperspectral data to improve FPAR estimating precision.展开更多
基金Project supported by the National Key Technology Research and Development Program of China (Nos.40801167 and 2006BAD05B05)the Knowledge Innovation Program of the Chinese Academy of Sciences (No.KZCX3-SW-356)the Foundation of the Chinese Academy of Sciences for the Field Stations of Resources and Environment
文摘Several studies have demonstrated that soil reflectance decreases with increasing soil moisture content, or increases when the soil moisture reaches a certain content; however, there are few analyses on the quantitative relationship between soil reflectance and its moisture, especially in the case of black soils in northeast China. A new moisture adjusting method was developed to obtain soil reflectance with a smaller moisture interval to describe the quantitative relationship between soil reflectance and moisture. For the soil samples with moisture contents ranging from air-dry to saturated, the changes in soil reflectance with soil moisture can be depicted using a cubic equation. Both moisture threshold (MT) and moisture inflexion (MI) of soil reflectance can also be determined by the equation. When the moisture range was smaller than MT, soil reflectance can be simulated with a linear model. However, for samples with different soil organic matter (OM), the parameters of the linear model varied regularly with the OM content. Based on their relationship, the soil moisture can be estimated from soil reflectance in the black soil region.
基金Many thanks are given to the Knowledge Innovation Program of the Chinese Academy of Sciences （KZCX-SW-356） the National Natural Science Foundation of China （40401003）, and the Project of Field Station on Resources and Environment of the Chinese Academy of Sciences.
文摘Fraction of absorbed photosynthetically active radiation （FPAR） is one of the important variables in many vegetation productivity and biomass estimation models. Therefore, it is significant to retrieve FPAR accurately for the improvement of model precision. On the basis of the field experiment, this article analyzed the correlations between corn canopy FPAR and spectral reflectance, and reflectance derivative. Discussion about the mechanism of FPAR estimation with different empirical models is based upon corn canopy reflectance, reflectance derivative, NDVI （normalized difference vegetation index） and RVI （ratio vegetation index）. The reflectance of visible bands showed much better correlations with FPAR than near-infrared bands. The correlation between FPAR and reflectance derivative varied more frequently and greatly than that between FPAR and reflectance, and with preferable correlation only around 520, 570, 670, 805, 950, and 1 010 nm. Reflectance and reflectance derivative both had intimate correlation with FPAR at some typical single band, with the maximum R^2 of 0.791 and 0.882, respectively. In a word, reflectance derivative and vegetation index were much effective in the estimation of corn FPAR than reflectance, and the stepwise regression of multibands with reflectance derivative showed the best regression with R^2 of 0.944. Reflectance at 375 and 950 nm with absorption characteristics caused by water showed prodigious potential for FPAR precisely estimating model establishment. On the whole, vegetation index and reflectance derivative had good relationships with FPAR, and could be used for FAPR estimation. It would be effective for choosing right bands and excavating the hyperspectral data to improve FPAR estimating precision.
