### 使用Python批量处理MODIS HDF文件 #### 读取HDF数据 为了有效地读取HDF格式的数据，可以利用`pyhdf`库来访问和操作这些文件。通过创建一个函数来遍历指定目录中的所有`.hdf`文件并逐一打开它们。 ```python import os from pyhdf.SD import SD, SDC def read_hdfs(directory_path): """Read all .hdf files from the given directory.""" data_sets = [] for filename in os.listdir(directory_path): if not filename.endswith('.hdf'): continue file_path = os.path.join(directory_path, filename) try: hdf_file = SD(file_path, SDC.READ) # Collect datasets within each HDF file. dataset_names = [k for k, v in hdf_file.attributes().items()] selected_datasets = {} for name in dataset_names: sds_obj = hdf_file.select(name) # select the dataset[^2] selected_datasets[name] = sds_obj.get() data_sets.append({ 'filename': filename, 'datasets': selected_datasets }) except Exception as e: print(f"Failed to process {file_path}: {str(e)}") raise return data_sets ``` #### 转换与保存为TIFF格式 对于将HDF数据转换成更易于使用的地理空间栅格图像(TIFF)，GDAL是一个强大的工具包。下面展示了一个简单的例子说明如何完成这一过程： ```python from osgeo import gdal, ogr, osr import numpy as np def convert_to_tiff(hdf_data, output_dir): """ Convert a single HDF entry into multiple TIFFs based on its internal datasets. Parameters: - hdf_data (dict): A dictionary containing metadata about an HDF file including its datasets. - output_dir (string): Path where generated tiffs should be saved. """ driver = gdal.GetDriverByName('GTiff') for ds_name, array in hdf_data['datasets'].items(): out_filename = f"{os.path.splitext(hdf_data['filename'])[0]}_{ds_name}.tif" out_filepath = os.path.join(output_dir, out_filename) # Create new GeoTIFF with dimensions matching input arrays dst_ds = driver.Create(out_filepath, array.shape[1], array.shape[0], 1, gdal.GDT_Float32) # Set geotransform parameters here according to your needs... geo_transform = (-180.0, 0.05, 0, 90.0, 0, -0.05) dst_ds.SetGeoTransform(geo_transform) # Define projection information; modify EPSG code accordingly spatial_reference = osr.SpatialReference() spatial_reference.ImportFromEPSG(4326) dst_ds.SetProjection(spatial_reference.ExportToWkt()) band = dst_ds.GetRasterBand(1) band.WriteArray(array.astype(np.float32)) band.FlushCache() del dst_ds ``` 上述代码片段展示了基本的工作流程：先定义好输入路径以及输出位置；接着调用之前编写的`read_hdfs()`方法加载所有的HDF资源；最后逐一对每一个数据集执行`tiff`转化逻辑[^4]。 #### 批量处理脚本 这里提供一段完整的批处理程序框架用于自动化整个工作流： ```python if __name__ == "__main__": input_directory = "/path/to/hdf/files/" output_directory = "/desired/output/path/" hdfs = read_hdfs(input_directory) for item in hdfs: convert_to_tiff(item, output_directory) ```