前言介绍
- 近期需要用到日本葵花8号卫星数据，用于相关研究，而通过官方提供的下载方法，难以针对性的下载所需要的数据类型，因此这里编写了一个针对葵花8号卫星L1级数据产品的脚本下载，主要实现两个功能：
- 1、自定义时间下载
- 2、选择任意的分辨率下载
完整代码，下载5km分辨率，30min一次的指定时间内的数据
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
# ================================================================================================
# Author: Jianpu | Affiliation: Hohai
# Email : 211311040008@hhu.edu.cn
# Last modified: 2023-04-04 12:28:06
# Filename: download_kuihua8_L1.py
# Description: 1.日本葵花8号卫星，L1级数据产品下载脚本
# 2.需要提前下载好对于葵花8卫星产品下载的安装包： lb_toolkits.tools
# 3.在官网注册获取下载的用户名和密码以及链接网址
# 4. 可以选择两个分辨率进行下载：5km和2km
# 5. 可以选择下载的频率：10min一次或者30min一次
# =================================================================================================
"""
import os
import sys
import datetime
import time
from lb_toolkits.tools import ftppro
class downloadH8(object):
def __init__(self, username, password):
self.ftp = ftppro(FTPHOST, username, password)
def search_ahi8_l1_netcdf(self, starttime, endtime=None, pattern=None, skip=False):
'''
下载葵花8号卫星L1 NetCDF数据文件
Parameters
----------
starttime : datetime
下载所需数据的起始时间
endtime : datetime
下载所需数据的起始时间
pattern: list, optional
模糊匹配参数
Returns
-------
list
下载的文件列表
'''
if endtime is None :
endtime = starttime
downfilelist = []
nowdate = starttime
while nowdate <= endtime :
# 拼接H8 ftp 目录
sourceRoot = os.path.join('/jma/netcdf', nowdate.strftime("%Y%m"), nowdate.strftime("%d"))
sourceRoot = sourceRoot.replace('\\','/')
# 获取文件列表
filelist = self.GetFileList(starttime, endtime, sourceRoot, pattern)
# filelist = [f for f in filelist if f.startswith('NC_H08_') and f.endswith('.06001_06001.nc')]
if len(filelist) == 0 :
nowdate += datetime.timedelta(days=1)
print('未匹配当前时间【%s】的文件' %(nowdate.strftime('%Y-%m-%d')))
continue
nowdate += datetime.timedelta(days=1)
downfilelist.extend(filelist)
return downfilelist
def GetFileList(self, starttime, endtime, srcpath, pattern=None):
''' 根据输入时间，匹配获取H8 L1数据文件名 '''
downfiles = []
srcpath = srcpath.replace('\\', '/')
filelist = self.ftp.listdir(srcpath)
filelist.sort()
for filename in filelist :
namelist = filename.split('_')
nowdate = datetime.datetime.strptime('%s %s' %(namelist[2], namelist[3]), '%Y%m%d %H%M')
if (nowdate < starttime) | (nowdate > endtime) :
continue
downflag = True
# 根据传入的匹配参数，匹配文件名中是否包含相应的字符串
if pattern is not None :
if isinstance(pattern, list) :
for item in pattern :
if item in filename :
downflag = True
# break
else:
downflag = False
break
elif isinstance(pattern, str) :
if pattern in filename :
downflag = True
else:
downflag = False
if downflag :
srcname = os.path.join(srcpath, filename)
srcname = srcname.replace('\\','/')
downfiles.append(srcname)
return downfiles
def download(self, outdir, srcfile, blocksize=1*1024, skip=False):
"""通过ftp接口下载H8 L1数据文件"""
if not os.path.exists(outdir):
os.makedirs(outdir)
print('成功创建路径：%s' %(outdir))
if isinstance(srcfile, list) :
count = len(srcfile)
for srcname in srcfile:
count -= 1
self._download(outdir, srcname, blocksize=blocksize, skip=skip, count=count+1)
elif isinstance(srcfile, str) :
self._download(outdir, srcfile, blocksize=blocksize, skip=skip)
def _download(self, outdir, srcname, blocksize=1*1024, skip=False, count=1):
print('='*100)
basename = os.path.basename(srcname)
dstname = os.path.join(outdir, basename)
if skip :
return srcname
if os.path.isfile(dstname) :
print('文件已存在，跳过下载>>【%s】' %(dstname))
return srcname
stime = time.time()
print(datetime.datetime.utcnow().strftime('%Y-%m-%d %H:%M:%S'),
'开始下载文件【%d】: %s'%(count, srcname))
if self.ftp.downloadFile(srcname, outdir, blocksize=blocksize):
print(datetime.datetime.utcnow().strftime('%Y-%m-%d %H:%M:%S'),
'成功下载文件【%s】:%s' %(count, dstname))
else:
print(datetime.datetime.utcnow().strftime('%Y-%m-%d %H:%M:%S'),
'下载文件失败【%s】:%s' %(count, dstname))
etime = time.time()
print('下载文件共用%.2f秒' %(etime - stime))
return srcname
def check_data_completeness(file_list, start_time, end_time):
expected_num_files = (end_time - start_time).days *48 + 48 # 48 show 30min/time; 144 show 10min/time
actual_num_files = len(file_list)
if actual_num_files == expected_num_files:
print("已经下载了全部数据。")
else:
print("有 %d 个数据文件缺失。" % (expected_num_files - actual_num_files))
expected_file_names = []
actual_file_names = []
for i in range(expected_num_files):
file_time = start_time + datetime.timedelta(minutes=i * 30)
file_name = "NC_H08_%s_R21_FLDK.06001_06001.nc" % (file_time.strftime("%Y%m%d_%H%M"))
expected_file_names.append(file_name)
for file_path in file_list:
file_name = os.path.basename(file_path)
actual_file_names.append(file_name)
missing_file_names = set(expected_file_names) - set(actual_file_names)
for missing_file_name in missing_file_names:
print("缺失文件：%s" % missing_file_name)
FTPHOST='ftp.ptree.jaxa.jp'
# create an instance of the downloadH8 class
h8_downloader = downloadH8('xxx', 'xxx')
#
# search for H8 files for a specific date
start_time = datetime.datetime(2017, 1, 1)
end_time = datetime.datetime(2017, 1,31,23, 59, 59)
file_list = h8_downloader.search_ahi8_l1_netcdf(start_time, end_time,pattern=['R21','02401_02401'])
# 选取每30分钟的数据文件名
selected_files = []
for file in file_list:
if file.endswith(".nc"):
if file[40:42] in ["00", "30"]:
selected_files.append(file)
# 打印选取的文件名
print(selected_files)
check_data_completeness(selected_files,start_time, end_time)
from tqdm import tqdm
for file in tqdm(selected_files):
try:
h8_downloader.download('/media/DATA/kuihua8_5km/', file)
except ValueError as e:
print(str(e))
os.remove(os.path.join('/media/DATA/kuihua8_5km/', os.path.basename(file)))
h8_downloader.download('/media/DATA/kuihua8_5km/', file)
脚本主要完成如下功能：
- 创建一个downloadH8类，用于FTP文件下载相关操作；
- 定义了search_ahi8_l1_netcdf函数，用于搜索指定时间范围内的葵花8号卫星L1 NetCDF数据文件，返回下载的文件列表；
- 定义了GetFileList函数，用于获取FTP服务器上指定路径下的文件列表，并根据输入时间和匹配参数筛选符合条件的文件进行返回；
搜寻数据和获取服务器文件的脚本如下：
def search_ahi8_l1_netcdf(self, starttime, endtime=None, pattern=None, skip=False):
'''
下载葵花8号卫星L1 NetCDF数据文件
Parameters
----------
starttime : datetime
下载所需数据的起始时间
endtime : datetime
下载所需数据的起始时间
pattern: list, optional
模糊匹配参数
Returns
-------
list
下载的文件列表
'''
if endtime is None :
endtime = starttime
downfilelist = []
nowdate = starttime
while nowdate <= endtime :
# 拼接H8 ftp 目录
sourceRoot = os.path.join('/jma/netcdf', nowdate.strftime("%Y%m"), nowdate.strftime("%d"))
sourceRoot = sourceRoot.replace('\\','/')
# 获取文件列表
filelist = self.GetFileList(starttime, endtime, sourceRoot, pattern)
# filelist = [f for f in filelist if f.startswith('NC_H08_') and f.endswith('.06001_06001.nc')]
if len(filelist) == 0 :
nowdate += datetime.timedelta(days=1)
print('未匹配当前时间【%s】的文件' %(nowdate.strftime('%Y-%m-%d')))
continue
nowdate += datetime.timedelta(days=1)
downfilelist.extend(filelist)
return downfilelist
def GetFileList(self, starttime, endtime, srcpath, pattern=None):
''' 根据输入时间，匹配获取H8 L1数据文件名 '''
downfiles = []
srcpath = srcpath.replace('\\', '/')
filelist = self.ftp.listdir(srcpath)
filelist.sort()
for filename in filelist :
namelist = filename.split('_')
nowdate = datetime.datetime.strptime('%s %s' %(namelist[2], namelist[3]), '%Y%m%d %H%M')
if (nowdate < starttime) | (nowdate > endtime) :
continue
downflag = True
# 根据传入的匹配参数，匹配文件名中是否包含相应的字符串
if pattern is not None :
if isinstance(pattern, list) :
for item in pattern :
if item in filename :
downflag = True
# break
else:
downflag = False
break
elif isinstance(pattern, str) :
if pattern in filename :
downflag = True
else:
downflag = False
if downflag :
srcname = os.path.join(srcpath, filename)
srcname = srcname.replace('\\','/')
downfiles.append(srcname)
return downfiles
- 定义了download和_download函数，用于将符合条件的文件下载到本地指定目录；
def download(self, outdir, srcfile, blocksize=1*1024, skip=False):
"""通过ftp接口下载H8 L1数据文件"""
if not os.path.exists(outdir):
os.makedirs(outdir)
print('成功创建路径：%s' %(outdir))
if isinstance(srcfile, list) :
count = len(srcfile)
for srcname in srcfile:
count -= 1
self._download(outdir, srcname, blocksize=blocksize, skip=skip, count=count+1)
elif isinstance(srcfile, str) :
self._download(outdir, srcfile, blocksize=blocksize, skip=skip)
def _download(self, outdir, srcname, blocksize=1*1024, skip=False, count=1):
print('='*100)
basename = os.path.basename(srcname)
dstname = os.path.join(outdir, basename)
if skip :
return srcname
if os.path.isfile(dstname) :
print('文件已存在，跳过下载>>【%s】' %(dstname))
return srcname
stime = time.time()
print(datetime.datetime.utcnow().strftime('%Y-%m-%d %H:%M:%S'),
'开始下载文件【%d】: %s'%(count, srcname))
if self.ftp.downloadFile(srcname, outdir, blocksize=blocksize):
print(datetime.datetime.utcnow().strftime('%Y-%m-%d %H:%M:%S'),
'成功下载文件【%s】:%s' %(count, dstname))
else:
print(datetime.datetime.utcnow().strftime('%Y-%m-%d %H:%M:%S'),
'下载文件失败【%s】:%s' %(count, dstname))
etime = time.time()
print('下载文件共用%.2f秒' %(etime - stime))
return srcname
此外，对于_download函数，还可以进一步修改，判断如果文件没有下载成功，则重新下载，这里也给出我改的一个函数：
def _download(self, outdir, srcname, blocksize=1*1024, skip=False, count=1):
print('=' * 100)
basename = os.path.basename(srcname)
dstname = os.path.join(outdir, basename)
if skip:
return srcname
if os.path.isfile(dstname):
print('文件已存在，跳过下载>>【%s】' % (dstname))
return srcname
stime = time.time()
print(datetime.datetime.utcnow().strftime('%Y-%m-%d %H:%M:%S'),
'开始下载文件【%d】: %s' % (count, srcname))
while True:
if self.ftp.downloadFile(srcname, outdir, blocksize=blocksize):
print(datetime.datetime.utcnow().strftime('%Y-%m-%d %H:%M:%S'),
'成功下载文件【%s】:%s' % (count, dstname))
break
else:
print(datetime.datetime.utcnow().strftime('%Y-%m-%d %H:%M:%S'),
'下载文件失败, 正在重新下载【%s】:%s' % (count, dstname))
etime = time.time()
print('下载文件共用%.2f秒' % (etime - stime))
return srcname
- 定义了check_data_completeness函数，用于检查是否存在数据文件缺失；
对于5km分辨率，10min一次的的：
def check_data_completeness(file_list, start_time, end_time):
expected_num_files = (end_time - start_time).days * 144 + 144
actual_num_files = len(file_list)
if actual_num_files == expected_num_files:
print("已经下载了全部数据。")
else:
print("有 %d 个数据文件缺失。" % (expected_num_files - actual_num_files))
expected_file_names = []
actual_file_names = []
for i in range(expected_num_files):
file_time = start_time + datetime.timedelta(minutes=i * 10)
file_name = "NC_H08_%s_R21_FLDK.02401_02401.nc" % (file_time.strftime("%Y%m%d_%H%M"))
expected_file_names.append(file_name)
for file_path in file_list:
file_name = os.path.basename(file_path)
actual_file_names.append(file_name)
missing_file_names = set(expected_file_names) - set(actual_file_names)
for missing_file_name in missing_file_names:
print("缺失文件：%s" % missing_file_name)
对于2km分辨率，30min一次的的：
def check_data_completeness(file_list, start_time, end_time):
expected_num_files = (end_time - start_time).days *48 + 48 # 48 show 30min/time; 144 show 10min/time
actual_num_files = len(file_list)
if actual_num_files == expected_num_files:
print("已经下载了全部数据。")
else:
print("有 %d 个数据文件缺失。" % (expected_num_files - actual_num_files))
expected_file_names = []
actual_file_names = []
for i in range(expected_num_files):
file_time = start_time + datetime.timedelta(minutes=i * 30)
file_name = "NC_H08_%s_R21_FLDK.06001_06001.nc" % (file_time.strftime("%Y%m%d_%H%M"))
expected_file_names.append(file_name)
for file_path in file_list:
file_name = os.path.basename(file_path)
actual_file_names.append(file_name)
missing_file_names = set(expected_file_names) - set(actual_file_names)
for missing_file_name in missing_file_names:
print("缺失文件：%s" % missing_file_name)
- 具体下载方法：
在主程序中，先调用search_ahi8_l1_netcdf函数搜索符合条件的文件列表，再通过循环调用download函数将文件下载到本地。
在具体的操作过程中，需要按照实际情况替换FTPHOST、username和password等相关参数，以保证正确的网络连接。同时，还需要根据需求修改脚本中的时间范围等参数，以便精确地搜索和下载文件。
下载示例
下面主要下载了了：2017.01.01-2017.01.31日之间，整个区域上，5km分辨率，30min一次的所有数据，如何看空间分辨率在后面有数据介绍
FTPHOST='ftp.ptree.jaxa.jp'
# create an instance of the downloadH8 class
h8_downloader = downloadH8('xxx', 'xxx')
## 2016 1440\0240 loss
# search for H8 files for a specific date
start_time = datetime.datetime(2017, 1, 1)
end_time = datetime.datetime(2017, 1,31,23, 59, 59)
file_list = h8_downloader.search_ahi8_l1_netcdf(start_time, end_time,pattern=['R21','02401_02401'])
# 选取每30分钟的数据文件名
selected_files = []
for file in file_list:
if file.endswith(".nc"):
if file[40:42] in ["00", "30"]:
selected_files.append(file)
# 打印选取的文件名
print(selected_files)
check_data_completeness(selected_files,start_time, end_time)
from tqdm import tqdm
for file in tqdm(selected_files):
try:
h8_downloader.download('/media/DATA/kuihua8_5km/', file)
except ValueError as e:
print(str(e))
os.remove(os.path.join('/media/DATA/kuihua8_5km/', os.path.basename(file)))
h8_downloader.download('/media/DATA/kuihua8_5km/', file)
如果你想要下载5km，10min一次的，只需要把选取每30分钟的数据文件名那段循环删掉，直接从file_list中进行下载即可
def check_data_completeness(file_list, start_time, end_time):
expected_num_files = (end_time - start_time).days *144 + 144 # 48 show 30min/time; 144 show 10min/time
actual_num_files = len(file_list)
if actual_num_files == expected_num_files:
print("已经下载了全部数据。")
else:
print("有 %d 个数据文件缺失。" % (expected_num_files - actual_num_files))
expected_file_names = []
actual_file_names = []
for i in range(expected_num_files):
file_time = start_time + datetime.timedelta(minutes=i * 10)
file_name = "NC_H08_%s_R21_FLDK.02401_02401.nc" % (file_time.strftime("%Y%m%d_%H%M"))
expected_file_names.append(file_name)
for file_path in file_list:
file_name = os.path.basename(file_path)
actual_file_names.append(file_name)
missing_file_names = set(expected_file_names) - set(actual_file_names)
for missing_file_name in missing_file_names:
print("缺失文件：%s" % missing_file_name)
FTPHOST='ftp.ptree.jaxa.jp'
# create an instance of the downloadH8 class
h8_downloader = downloadH8('xxx', 'xxx')
## 2016 1440\0240 loss
# search for H8 files for a specific date
start_time = datetime.datetime(2017, 1, 1)
end_time = datetime.datetime(2017, 1,31,23, 59, 59)
file_list = h8_downloader.search_ahi8_l1_netcdf(start_time, end_time,pattern=['R21','02401_02401'])
# 打印选取的文件名
print(file_list)
check_data_completeness(selected_files,start_time, end_time)
from tqdm import tqdm
for file in tqdm(file_list):
try:
h8_downloader.download('/media/DATA/kuihua8_5km/', file)
except ValueError as e:
print(str(e))
os.remove(os.path.join('/media/DATA/kuihua8_5km/', os.path.basename(file)))
h8_downloader.download('/media/DATA/kuihua8_5km/', file)
数据介绍
这里，对于数据名称做简单介绍：
NC_H08_20170101_0100_R21_FLDK.06001_06001.nc
NC_H08_20170101_0030_R21_FLDK.02401_02401.n
NC_Hnn_YYYYMDD_hhmm_Rbb_FLDK.xxxxx_yyyyy.nc
一般下载下来的数据名称如上所示：
NC_H08 ： nn=08代表葵花8号，nn=09表示葵花9号
YYYYMDD： 年月日
hhmm： 小时分钟
Rbb：Rbb = R21，表示的是整个区域，波段从 “01” 到 “16”；Rbb=R14,表示的是日本区域，波段为14
xxxxx: (“2401”: 5km resolution,
“6001”: 2km resolution, )
yyyyy: (“2401”: 5km resolution,
“6001”: 2km resolution, )
参考
https://mp.weixin.qq.com/s/b_BlfXYTTY0oAnb7U0tCOQ
https://mp.weixin.qq.com/s/EFT8D7ElvlcTBXQ_LkAbRw
