Forest canopy height is one of important input parameters of ecosystem model,and had a high correlation with the estimation of biomass and the study of carbon cycle.This paper reviewed systematically the development history and the newest trends of forest canopy height studies,summarized the mainly models and methods of forest canopy height studies among different sensors(SAR tree height and LiDAR forest canopy height),including polarization interference model and tree height algorithm,ploarization coherence tomography(PCT),LiDAR canopy height estimation methods and forest canopy height profile simulation,deliberated on the intrinsic characteristics of different research techniques on the single sensor of forest canopy height,and analyzes the advantages and disadvantages of the multi\|sensor joint inversion method of the regional forest canopy height.Analysis shows that the interpolation results of Kriging and CoKriging methods is not very good,hybrid model is more suitable for the space interpolation of the regional forest height compared with other single time or space model.From the scientific trends and social needs,the existing problems and difficulties were understudied and the new challenge and chances in forest canopy height studies were analyzed.This will provide new trajectories and new methods for further studies of forest vertical structure and global carbon cycle.
Noise in spectral dimension makes the spectrum distorted or deformed,shifting the central wavelength,affecting the precision of extracting land cover information and inverting surface parameters.Therefore,spectral denoising is of great significance for improving the effects of quantitative applications of hyperspectral remote sensing.Because the principle of mathematical morphological filtering is simple and easy to implement,it has been used in the study of vegetation spectra and fluorescence spectra of organic compounds.Mathematical morphology filtering was used to remove spectral noise of wheat.Intuitive analysis was made according to the spectrum similarity and spectral denoising effects;Moreover,the quantitative evaluation in practical application was made by adopting different vegetation indices to invert the biophysical and biochemical parameters.The results show that,compared with the traditional Savitzky\|Golay filter,the mathematical morphology filtering can keep the inherent characteristics in visible and near\|infrared region,and improve the accuracy of inverting Leaf Area Index and Chlorophyll slightly,which is caused mainly by the low spectral noise in the range of the spectrum.The mathematical morphology filtering can remove the noise in shortwave infrared (SWIR) region effectively,improving the accuracy of inverting foliar water content of wheat.But the traditional Savitzky\|Golay filter can only weaken the large scale noise in SWIR region.After generalized morphology filtering,the determination coefficient of regression (R2) between the vegetation index and foliar water content can reach 0.5130,while R2 is 0.3753 before filtering;R2 between inverted value and the measurement of foliar water content can reach 0.4221 with a root mean square error (RMSE) of 0.0243,while the corresponding values are 0.3097 and 0.0318 for R2 and RMSE before filtering.The results of the mathematical morphology filter are better than those of the traditional Savitzky\|Golay filter.
Aerodynamic roughness is a very important parameter to represent the aerodynamic characters in urban areas.Radar remote sensing is considered to be an effective means for aerodynamic roughness retrieval.This paper analyzed the direction features and scale characteristics of backscattering coefficient in northern Beijing by using 22 ALOSPALSAR images from 2006 to 2011.And aerodynamic roughness was calculated by the gradient data of wind speed and direction.The relationship between aerodynamic roughness and backscattering coefficient was analyzed in different scales and orientations.The results showed that the correlation coefficient was reached maximum when the radius of windward sector domain was 2 500 m and included angle was 30°.It indicated that the SAR images can effectively characterize the aerodynamic characters.The resultsprovidedimportant basis for the inversion of aerodynamic roughness by using SAR images,which will provide more accurate parameters for regional climate model and atmospheric boundary layer model.
In order to study spatial scale effect of lake water body extraction method based on remote sensing imagery,Landsat TM images are first upscaled to 100,250,500,1 000,2 000,3 000,4 000 and 5 000 m spatial resolutions from the initial 30 m image.Then Mixture Tuned Matched Filtering(MTMF)method is selected to conduct sub pixels unmixing and threshold method is applied to segment matched filtering score image to extract lake areas.Lastly,the spatial scale effect of remote sensing water body extraction is analyzed for different lakes with different sizes.The results lead to the following conclusions: ①the nearest neighbor resampling method is proved to be more suitable than the pixel aggregate method for scale upscaling of remote sensing images with higher spatial resolution in spatial scale effect study on lake areas. ②MTMF is a good choice for lake area extraction from remote sensing images at different spatial resolutions.But it is difficult to separate rivers or other water bodies from lakes. ③The lake areas retrieved from upscaled remote sensing images vary with different spatial resolutions of these images.For one specific lake,the coarser the remote sensing image is,the larger overestimations of the lake area will be obtained,especially for lakes with small area.This work is very important for lake water body mapping with remote sensing at different spatial resolutions.It can not only provide useful information to understand the difference between lake areas retrieved from different remote sensing data,but also can help researchers to choose suitable reference data to validate their remote sensing lake areas.It would be very useful for remote sensing lake areas application in regional or global environment and climate change study.
It is a fast and nondestructive way to detect the foliar biochemistry of tea using reflectance spectroscopy.To analyze the relationship between the spectra and the contents of foliar biochemistry,the spectra of young leaves,adult leaves and old leaves from four kinds of tea was measured,and corresponding contents of chlorophyll,total tea polyphenols and amino acids were obtained firstly.Then sensitive spectrum wave bands were determined by Optimum Index Factor and linear models were established to predict the concentrations of the foliar biochemistry by multiple regression.At last,differences of three spectral preprocessing methods (multiple scattering correction,standard normal variation,Savitzky\|Golay) in model building were discussed,and probable causes of the sensitive waves were analyzed.The results of this paper indicated that the accuracy was the highest when estimating the contents of chlorophyll,R2 could be as high as 0.9 between the estimated and measured contents of chlorophyll.Followed by the accuracy of theanine,R2 was about 0.7 between the estimated and measured contents.However,R2 was the lowest between the estimated and measured contents of polyphenols,only about 0.65.
The synchronous monitoring for the cyanobacteria bloom and aquatic plants is of great significance for the study of lake water environment、ecology and the water cycle.Compared with traditional monitoring methods,for instance,field investigation,using remote sensing technology with the advantages of large scope,long cycle,high efficiency and low cost.A model based on “Chlorophyll a Spectral Index” and “Baseline of Phycocyanin” was built to distinguish cyanobacteria bloom and aquatic plants in Lake Taihu by using Hyperspectral Imager for the Coastal Ocean (HICO) images.The average accuracy of cyanobacteria bloom and aquatic plants are 93% and 95% respectively.By overlapping the distribution maps of cyanobacteria bloom and aquatic plants,the distribution rules of cyanobacteria bloom and aquatic plants in Lake Taihu from 2010 to 2014 were analyzed,which are consistent with the former results in the literatures.The average thresholds were used to extract cyanobacteria bloom and aquatic plants,and the accuracy are 75.7% and 84.0% respectively.If the efficiency is more desired than accuracy,then average thresholds can be used to extract cyanobacteria bloom and aquatic plants.Which is convenient for realizing batch processing and the automation extraction of cyanobacteria bloom and aquatic plants.
The traditional Moderate Resolution Imaging Spectroradiometer(MODIS) fire detection algorithm relies primarily on hot spot detection using brightness temperature data derived from the 4 and 11 channels.There are limitations to the effectiveness of this algorithm when it is applied to monitor forest fires in different regions and four seasons.In response to these problems,a detailed description of an improved algorithm based on brightness temperature,vegetation index and Aerosol Optical Depth (AOD) is offered by adjusting the corresponding potential fire thresholds and contextual thresholds in the traditional algorithm,by exploring the differences in the post\|fire and pre\|fire values of the Normalized Difference Vegetation Index(NDVI),and by analysing the obvious increase in the AOD on the leeward side caused by the spread of a smoke plume.This approach is confirmed by several fire events in China.The study reveals that the improved algorithm achieves significantly lower false alarm rates and is more sensitive to cool fires.Then the adaptability of this algorithm in all environment is also enhanced.
To improve the precision of the satellite altimeter wind speed and sea state bias products,it is necessary to calibrate the backscatter coefficient measured by the radar altimeter.Distributed targets are good candidates in satellite backscatter calibration,and deserts which have constant backscatter characteristic are pretty suitable for the calibration of nadir-looking radar altimeters.In this paper the echo waveforms of Jason-2 altimeter from desert regions were checked and proper areas in the Simpson Desert,Australia and Takla Makan Desert,Xinjiang,China were chosen.Jason-2 altimeter data of seven years were collected in these areas.Afterwards,the characteristic of the altimeter backscatter coefficient over desert surface were investigated based on the data prepared,and a method of backscatter coefficient calibration based on the desert areas was proposed.The principle,feasibility and experiment plan were described,the key technologies such as the matching of the altimeter and calibrating scatterometer footprint were discussed,and the calibration errors were analyzed.It was shown that a 0.5 dB calibration precision specification could be achieved.The method proposed in this paper would significantly contribute in the HY-2A mission by improving the qualities of wind speed and sea state bias products.
To monitor a space object in ISAR (Inverse Synthetic Aperture Radar) technology,how to extract 3-D information from 2-D imaging is an interesting challenge in ATR (Automatic Target Recognition).A numerical approach,so\|called the Bidirectional Analytic Ray Tracing (BART) method,is first employed to numerically calculate the scattering of a complex object in space,e.g.an Envisat model under different angular positions.And then,the sequence of 2\|D ISAR imaging with angular positions is produced.Kanade-Lucas-Tomasi (KLT) feature tracker is adopted for feature extraction and matching among consecutive ISAR images.3-D positions of featured points are estimated by the orthographic factorization method (OFM).The ISAR image sequence of a simple hexagonal frustum model is simulated firstly.Then,3-D position of corner points are estimated and the result shows good accuracy.Simulated ISAR image sequence of Envisat satellite is taken as an example.The results show that 3-D shape information can be properly acquired from 2-D ISAR image sequence.
A new model is proposed to estimate the significant wave height and average period with ERS-1/2 scatterometer data.The relationship between the wind speed and significant wave height in full developing wind\|wave domination is established by Russian scholars.Through this formula,the data that based on the ERS1/2 data and the NDBC buoys data matching were distinguished into three states:developing wind\|wave,full developing wind\|wave and swell wave,respectively.The significant wave heights and average period are retrieved by Back Propagation neural network,the root mean square is 0.53,0.57,0.90 m and 0.69 s,1.04 s,1.36 s in three states,respectively.This method inversion significant wave height and period is found that the full developing wind\|wave domination is best effect,in turn,is the developing wind\|wave,the last is swell wave.The study provides the basis for scatterometer data inversion of wave parameters.It makes the biggest possible inversion wave parameters in large area.
High precision coastal altimetry by using satellite radar altimeter faces many problems,re-construction of satellite radar altimeter’s coastal return waveforms are contaminated by the return waveforms from the land and building satellite radar altimeter measurements datum over coastal are key to improve accuracy of coastal altimetry.Introducing a method of re-construct satellite radar altimeter measurements datum over coastal waters based on transponder,taking HY-2 satellite radar altimeter as an example,respectively analyze satellite radar altimeter’s coastal return waveforms are contaminated in coastal waters of more than 20km offshore distance,less than 2km offshore distance and 2~20 km offshore distance,then induce the accuracy of building satellite radar altimeter measurements datum over coastal waters in the theory.The HY-2 radar altimeter’s return waveforms data acquired from July 5,2012 in Tangshan’s coastal waters in the in-orbit calibration test,firstly observe HY-2 radar altimeter’s return waveforms forworded by transponder over coastal waters,the results show that:the feasibility of the method presented in this paper is verified,the method will help to improve accuracy of coastal altimetry.
Long time series of earth observation data records play an important role in the research of earth environment changes and trends.Few of today’s satellite sensors can continually work for more than ten years.So it becomes much necessary to conduct inter\|calibration of sensors with similar physical configuration and over-lapped observation.However,the different system parameters and process of calibration make difference even during the concurrent observation in the same zone.AMSR-E and AMSR2 share the similar configuration and it is possible for us to select the overlapped observation during the period of 2013~2014 at 18.7 GHz and 36.5 GHz.After comparing the data,a least-square linear model was established to calibrate the two sensors.The results show that correlation between them up to 0.98 and the value difference ranges from 3 K to 6 K.This difference should not be ignored when constructing a long-term data record or during the retrieval of the land surface parameters.
Since 1990s,the satellite altimeter made a great contribution to the monitoring of sea level changes,ocean circulation,regional sea surface topography,climate change et.al.However,satellite altimeter has bias or drift when measuring the Sea Surface Height(SSH),so it needed to be calibrated.The principle of calibration is to measure the sea surface height in the same position of satellite when it travel through,then compare the in\|situ SSH with the value of altimeter measurements,and achieve the bias or drift of the altimeter.In this paper,a self-designed GPS buoys was used to measure the SSH when the satellite transit above.Four experiments was done to calibrate HY-2 altimeter in Shidao town of Shandong province and Qinglan town of Hainan province.The results showed that:①In support of GPS reference station,the precision of GPS buoy in measuring the SSH was better than 50mm.This could satisfy the demand of altimeter calibration.②There was a systematic bias in HY-2 satellite altimeter measurements.③By using the data of relative calibration between Jason\|2 and HY\|2 altimeter and absolute calibration results,HY-2 altimeter can reach the international advanced level.
High spatial resolution remote sensing images provide an important and effective way to study and analyze human activities on finer scales.However,the diversity and complexity of buildings make it a challenging task to efficiently and accurately extract building information from high resolution images.This paper proposed a method named Multi-structural element of Morphological Building Index(MMBI) to extract building information.Based on variety shapes and structural characteristics of buildings presented in the high-resolution images,the multi\|structural morphological building index is developed,by combining with guided filter which performs well in denoising and preserving edges.The proposed method was compared with the MBI approach through analyzing their performances in extracting buildings with different characteristics from a GeoEye-1 image.Results show that the MMBI has an accuracy rate above 88.2%,which is 5% higher than that of the MBI.Meanwhile,the omission rate of the MMBI method is 6% lower than the MBI.The results proved that the MMBI method is able to quickly and accurately extracting building information from high spatial resolution remote sensing images and thus has great potential of application in automatic building extraction.
Due to the aerosol and other reasons,large area of remote-sensing chlorophyll data are often missing,which hinders its development,and also becomes a hot topic in this field.Therefore,to retrieve the long\|term synchronous data,the re\|constructure and re\|analysis of remote-sensing data is called for.This can extend the application of remote-sensing data,and improve its utilization efficiency.In this study,combining the advantages of the DINEOF method and the Optimal Interpolation (OI),an algorithm (DINEOF-OI) with multi\|variables and second\|order\|correction has been developed to study the variable chlorophyll distribution in multi scales including weather-process scales in the East China Sea (ECS).This paper introduces how to locate the influencing weights of observed data on re-constructed data in temporal and spatial series.based on the algorithm developed in this paper,the recent decadal chlorophyll data in the ECS has been re-constructed and analysed.It indicates the re-constructed database can describe the primary characteristics of chlorophyll distribution in the ECS in multi\|scale processes.
A fused image using multispectral bands of Landsat-8 OLI and panchromatic band of GF-1 PMS is used to extract iron stained alteration information in the study area.The method of principal component analysis is used in this study.In order to verify the advantage of the fused data in mineralized information extraction,a comparison of the extracted iron stained alteration information between before and after data fusing is done.The result shows that it can extract more alteration information by using the fused data than the non-fused data,especially for small features.
Multi\|source data and various technological methods are used in the study of vegetation phenology.But the differences of various phenological methods based on multi-source data still need to be explored.We initiated research on winter wheat field ecosystem northern China in Yucheng City of Shandong Province.Here we explored the differences of the main phenological phases of winter wheat obtained from multi\|source data:normalized difference vegetation index,enhanced vegetation index,digital camera,carbon flux data and field-measured data.The main phenological phases of winter wheat are the date of green-up,maturity,and the length of growing season.We found the phenological phases of winter wheat obtained from carbon flux data are the nearest to the result of field-measured,and the differences of various periods are less than 3d.The results obtained from the digital camera are poor than the carbon flux data,and the remote sensing data (NDVI and EVI) are the worst.The results of phenological phases of winter wheat obtained by NDVI and EVI have extremely significant correlation,and the period of green-up has the strongest correlation (R2=0.857,P<0.001).All the results of growth period of winter wheat based on multi\|source data showed that the dates of green-up are advanced,maturity are delayed,and the lengths of growing season are longer than before.
Studying vegetation phenology and the relationship between climate and phenology is of vital importance in understanding the change of global ecological environment.Near-surface digital camera has become an effective detection method in vegetation phenology due to its frequent and accurate data\|acquirement capability.Taking Vaira Ranch in North America for example,this study used digital camera based time series of Greenness Chromaticity Coordinates (Gcc) to monitor the growing condition in spring and to model its phonology.The phonological metrics were extracted from Gcc,after which GPP and meteorological data were used to compare and analyze with the extracted metrics.The results showed that vegetation within research area grew from the 20th to 145 th day of year,and the correlation index between Gcc and GPP was 0.88,the relative mean deviation of phonological features was 0.05.Meanwhile,precipitation,soil moisture,soil temperature and solar radiation are factors that can affect the vegetation growth.In details,the combination of atmosphere temperature,soil temperature and solar radiation can explain 91.3% of Gcc change,among which,atmosphere and soil temperature can explain the most of phonology with 30.9% and 49.0% individually.Furthermore,as the lacking of water content,precipitation behaves as an essential factor for inhibiting vegetation growth in our research area.
Only using the post\|earthquake PolSAR imagery to interpret collapsed buildings information is a rapid and effective disaster investigation means.,but also is easy and fast for implementation.In PolSAR images,collapsed buildings and oblique undamaged buildings of which the orientation is unparallel to the radar flight direction all present volume scattering characteristics,and the over\|classification of collapsed buildings and the under-classification of undamaged buildings will be generated,which can lead to the excessive evaluation for disaster losses.Because of the difference of the significant texture features of the collapsed buildings and the oblique buildings,the difference of texture will be used to solve the problem of the mixing of the collapsed building and the oblique building.After research and analysis,collapsed buildings and oblique buildings can be well distinguished by their differences in the four texture feature parameters of Mean,Homogeneity,Entropy,and Correlation based on Gray-Level Co-occurrence Matrix (GLCM).Therefore,in order to reduce the false alarm rate of collapsed buildings,the four kinds of texture feature will be used to extract the oblique buildings and collapsed buildings.In this work,the Yushu earthquake was taken as example,and building earthquake damage information in urban region were extracted.The experimental results show that the proposed method can greatly increase the building earthquake damage assessment accuracy.
It is a problem that how to eliminate the interference of mountain shadows when we extract the water body in the mountainous areas.The spectrum\|based methods such as spectral correlation or spectral threshold controlling may partially remove the shadows.But the real conditions of water body mapping were rather complex,and the above\|mentioned methods may underestimate or overestimate the lake water extents based on different thresholds,even missing some lakes.The method proposed in this study firstly uses the spectral feature of water to extract the potential complete lakes,then remove the shadows by differentiating the relationships of spatial position.As a result,we cannot only keep relative complete lakes,but also reduce the interference of mountain shadows efficiently.By adjusting the parameter for minimal size of lake extraction,we can remove more than 90% of the mountain shadows on the premise of saving relatively complete lakes in the high mountain areas.
Pixels and patches allocated by supervised classification are often very scattered and cluttered.To make these classified patches more complete to form image objects and with only tolerable errors,a new algorithm,named Self\|adaptive Generalization of Classified Patch (SGCP),has been proposed.It uses some typical algorithms of both image analysis and map generalization,and has sure progress in the combined applications of the two.By using SGCP,it has been achieved to conduct generalization of classified patches of land use in a self\|adaptive way.Meanwhile,a better balance between degree and precision of the generalization can be promised.
There are six steps to conduct SGCP:①performing a supervised classification in a multi\|descriptor space;②separating road from the rest patches of impervious surface via binary morphology operations and by using shape descriptors;③removing noises and making these patches more complete via binary morphology operations and by filtering out smaller patches;④simplifying the boundary of a patch by recursively backfilling its convex residuals into a convex hill of the patch;⑤eliminating gaps between patches by merging the gaps into surrounding larger patches and ⑥ assessing the generalization degree of a patch by checking the vertex reduction rate of a convex hill of the patch and,in the same time,assessing the generalization accuracy by checking whether the area of each class and the global accuracy of all the classes are maintained as well as possible.Some main parameters,such as size of structure element,recursion times,neighboring size etc.,are self\|adaptively determined.In addition,there are several reserved parameters to allow the degree of generalization adjustable by user as the user is unsatisfied with the results of automatic generalization.Simulation tests with Matlab show that by using the algorithm proposed in this paper,a proper balance between the degree and accuracy of generalization can often be guaranteed.An increase of 22.9% in the generalization degree will lead only a decrease of 0.72% in the global accuracy and a change of 2.72% in the patch area.
A smoothing algorithm based on discrete wavelet was firstly applied to the MODIS EVI time\|series data from 2001 to 2012 to minimize the effects of anomalous values caused by atmospheric haze and cloud contamination,based on dynamic threshold method to extract vegetation phenology information of China,study the spatio\|temporal variation characteristics of crops and natural vegetation phenology,The results indicate:①In the first season of crops,start date,peak date and end date mainly with altitude rise delayed centred on north China Natural vegetation phenology advance than crops about 20 days,As the rise of the altitude to delay first,and then in advance;②In temporal,Phenological have significant changes area of the first quarter,43.98% of start date,52.83% of peak date have trends in advance,Mainly in southwest,the junction of the northeast and Inner Mongolia,where the start date later and end date early.The rest of the region,start date,peak date and 81.80% of end date showed a trend of delay,Mainly in Loess Plateau and double season crop area;Crop phenological delay rate is less than natural vegetation.
With the high spatial resolution of optical data and the lack of weather effects of passive microwave data,we developed a daily cloud\|free snow cover product with a resolution of 500 m for region in Eurasia by the use of the Moderate Resolution Imaging Spectroradiometer (MODIS) standard daily snow cover products and the Advanced Microwave Scanning Radiometer\|Earth Observing System (AMSR\|E) snow water equivalent (SWE) product.Selecting Landsat\|TM data with high\|resolution as“truth value”images,the accuracy of the MODIS daily cloud\|free snow products over Eurasia is evaluated by comparing with these“truth value”images.The results demonstrated that MOD10A1 and MYD10A1 cannot be directly used to monitor snow cover because of the serious effect of cloud amount.However,the products developed in this study has good accuracy,which are high concordant with Landsat\|TM.Land\|use type has influence on snow classification accuracy.The accuracies of barren and grasslands are best and the Kappa coefficient reaches to 0.655 and 0.644 respectively.The accuracies of shrublands and croplands are good and the Kappa is 0.584 and 0.572,respectively.But the Kappa of the forested region is only 0.389.The integral accuracy of the synthesized product whose Kappa reaches to 0.569 in this study is high,which can effectively improve the accuracy of snow area monitoring in Eurasia.
The Mogolian Plateau (MP) is one of the typical arid regions around the world.The ecosystem of this region and its change have a lot to the areas surrounding the Mogolian Plateau.In particular,the vegetation system of the Mongolian Plateau plays an important ecological role in northern China and the whole of Northeast Asia.Thus,it is urgent to do research on the vegetation system to see how it has changed throughout these years and detect the trend.In this study,we investigate the spatial\|temporal pattern,processes and driver,and the degree and mechanism of grassland dynamics in the MP using the NOAA/AVHRR 10\|day maximum NDVI composite data of 1982~1999 and MODIS 16 day maximum NDVI data of 2000~2013 by building a conversion relationship of them.based on time series analysis methods,we calculate the slope of the fitted line of NDVI for each single grid cell,which is set as the indicator of the intensity of dynamic changes.Combined with meteorological data,we use the correlation analysis to explore how the natural factors affect the ecosystem.The results show that the temperature,precipitation and the vegetation are characterizing of zonality in Mongolia Plateau.Non\|vegetated area (NDVI<0.1) and Sparsely vegetated area (0.1<NDVI<0.2) account for 7.1% and 22.8% respectively;NDVI significant(p=0.05) increases about 27.0% of this area and significant decreases 7.3% of this area,which indicates the improvement of the vegetation system;the average CV of MP is 16.99%,which means the vegetation cover changes a lot;the area with significant positive correlation of NDVI and precipitation of growing season account for 9.2% in grassland of the semi\|arid area,and significant negative of NDVI and average temperature of growing season account for 14.8%,mainly in the north of MP.
