Moss moisture content mapping, Casey Station region, Antarctica

Data was collected by Ryan Haynes and Emiliano Cimoli, University of TasmaniaProcessed by Darren Turner, University of TasmaniaFor each site a DSM (Digital Surface Model) and Orthophoto was created based on images collected with a DJI Phantom 4 drone equipped with an RGB digital camera. These products were georeferenced based on ground control points placed within the study site for which their positions were accurately measured with a differential GNSS to within 2-4 cm accuracy.Sites were then flown with two hyperspectral pushbroom scanners, Specim FX10 and FX17 carried on board a DJI Matrice 600 drone. Pushbroom data was synchronised with onboard GNSS and IMU (inertial measurement unit) data to allow the scan lines to be projected onto the previously collected DSM, providing accurately georectified hyperspectral imagery. Hyperspectral datasets were than validated against the RGB orthophotos to confirm accurate georectification.Three sites containing moss were imaged using this workflow during February 2022, Robinsons Ridge, ASPA135 Lake, and ASPA135 Pond, details of coverage is found in he metadata.The purpose of collection the hyperspectral imager was to develop a model for predicting moss moisture content based on hyperspectral imagery. A model was developed in the lab to relate the spectral signature of a moss sample at various stages of drying to the hyperspectral signature of the moss. This model was then applied to the the hyperspectral imagery from each site such that a map of moss moisture could be created. Summary of the imagery collected at each site.DSM - raster file with each point representing the elevation of the ground surface in metres Orthophoto - georectified orthomosaic of the site where each pixel has an RGB value, such that is appears as it would to the human eye. FX10 hyperspectral - a georectified map made up of several passes with the pushbroom hyperspectral camera covering 224 bands between 397nm and 1003nm, other known as the Visible Near Infrared (VNIR), individual calibrated pixel values floating point values scaled between 0 and 1, representing 0% - 100% reflectance. Hyperspectral files are stored in band interleaved by pixel and have an accompanying header file in ENVI format with all required information. FX17 hyperspectral - a georectified map made up of several passes with the pushbroom hyperspectral camera covering 112 bands between 937nm and 1718nm, other known as the Short Wave Infrared (SWIR) individual calibrated pixel values floating point values scaled between 0 and 1, representing 0% - 100% reflectance. Hyperspectral files are stored in band interleaved by pixel and have an accompanying header file in ENVI format with all required information. Moss moisture maps - a georectified map of derived moss moisture based on the model developed in the laboratory. Individual calibrated pixel values floating point values scaled between 0 and 1, representing 0% - 100% moisture content, where 100% would be saturated moss.

本数据集由塔斯马尼亚大学的Ryan Haynes与Emiliano Cimoli采集，经塔斯马尼亚大学的Darren Turner处理。针对每个研究站点，研究者使用搭载RGB数码相机的DJI Phantom 4无人机采集影像，据此生成数字表面模型（DSM，Digital Surface Model）与正射影像。上述成果基于研究站点内布设的地面控制点完成地理配准，控制点位置通过差分GNSS精准测量，定位精度达2-4厘米。

随后，研究者使用搭载于DJI Matrice 600无人机的两台推扫式高光谱扫描仪——Specim FX10与FX17，对各站点开展航飞作业。将推扫式扫描数据与机载GNSS、惯性测量单元（IMU，inertial measurement unit）数据进行同步，以便将扫描线投影至此前采集的DSM上，生成经过精准地理校正的高光谱影像。后续通过RGB正射影像对高光谱数据集开展验证，确认其地理校正精度符合要求。

2022年2月，研究者采用该工作流程对三个包含苔藓的研究站点完成成像作业，分别为Robinsons Ridge、ASPA135湖与ASPA135池塘，各站点的覆盖范围详细信息可查阅元数据。

本次采集高光谱影像的核心目标为构建基于高光谱影像的苔藓含水量预测模型。研究者在实验室中建立模型，将不同干燥阶段的苔藓样本光谱特征与苔藓的高光谱特征建立关联。随后将该模型应用至各站点的高光谱影像，进而生成苔藓含水量空间分布图。

各采集影像的汇总说明如下：
1. 数字表面模型：栅格格式文件，每个栅格点代表地面高程，单位为米。
2. 正射影像：经过地理校正的站点正射镶嵌图，每个像素携带RGB色彩值，成像效果与人类肉眼直接观测一致。
3. FX10型高光谱数据：由推扫式高光谱相机多次航飞拼接生成的地理校正影像，涵盖397nm至1003nm波段范围内的224个波段，即可见光近红外波段（VNIR，Visible Near Infrared）。经校准的单个像素值为0至1之间的浮点型数值，分别对应0%至100%的反射率。高光谱文件采用按像素波段交错（band interleaved by pixel）的格式存储，并附带ENVI格式的头文件，包含所有必要元数据。
4. FX17型高光谱数据：由推扫式高光谱相机多次航飞拼接生成的地理校正影像，涵盖937nm至1718nm波段范围内的112个波段，即短波红外波段（SWIR，Short Wave Infrared）。经校准的单个像素值为0至1之间的浮点型数值，分别对应0%至100%的反射率。高光谱文件采用按像素波段交错的格式存储，并附带ENVI格式的头文件，包含所有必要元数据。
5. 苔藓含水量分布图：基于实验室构建的模型生成的地理校正苔藓含水量空间分布图。经校准的单个像素值为0至1之间的浮点型数值，分别对应0%至100%的含水量，其中100%代表完全饱和的苔藓。