批量TIF转NC并进行MK检验
这里主要记录一个批量进行tif文件转nc，并且将长序列数据进行mk检验的python代码。有问题随时联系：jia5678912。
python
import os
import numpy as np
import xarray as xr
from osgeo import gdal, osr
def Search_File(dirname,suffix):
'''
This function can search all files with the specified suffix in this dir.
:param dirname: string, the path need to be searched
:param suffix: string, the specified suffix need to be seached
:return:
filter_list: list, the path list need to be searched.
'''
filter = [suffix] # 设置过滤后的文件类型 当然可以设置多个类型
filter_list = []
for maindir, subdir, file_name_list in os.walk(dirname):
#print(maindir) #当前主目录
for filename in file_name_list:
apath = os.path.join(maindir, filename)#合并成一个完整路径
portion = os.path.splitext(apath)
ext = portion[1] # 获取文件后缀 [0]获取的是除了文件名以外的内容
if ext in filter:
newname = portion[0] + suffix
filter_list.append((newname,portion[0].split("\\")[-1]))
# print(filter_list)
return filter_list
def combin():
'''
数据合并，沿时间轴
:return:
'''
import xarray as xr
import pandas as pd
import os
# 假设你的NC文件存放在'data_folder'目录下
data_folder = r"E:\530\VQI\vqi" # 请替换为你的NC文件所在的文件夹路径
file_n = Search_File(data_folder,".nc")
nc_files = [p for p,n in file_n]
# n = [n for p,n in file_n]
print(nc_files)
# nc_files = [os.path.join(data_folder, f) for f in os.listdir(data_folder) if f.endswith('.nc')]
first_file = xr.open_dataset(nc_files[0])
x = first_file["lat"].values
y = first_file["lon"].values
change_file_p = r"{}\change".format(data_folder)
if not os.path.exists(change_file_p):
os.makedirs(change_file_p)
for file,n in file_n:
# file_path = os.path.join(folder_path, file)
ds = xr.open_dataset(file)
ds['lat'] = x
ds['lon'] = y
ds.to_netcdf(r"{}/{}.nc".format(change_file_p,n))
ds.close()
nc_files_new = [os.path.join(change_file_p, f) for f in os.listdir(change_file_p) if f.endswith('.nc')]
# 排序文件保证时间顺序
nc_files_new.sort()
# 使用open_mfdataset来读取所有的.nc文件，假设所有文件都有相同的x和y坐标
combined_ds = xr.open_mfdataset(nc_files_new, concat_dim='time', combine='nested')
# 创建时间坐标（这里我们从文件名中提取时间信息，假设格式为YYYY.*）
print(nc_files_new)
time_coords = [pd.to_datetime(f.split('.')[0][-4:], format='%Y') for f in nc_files_new]
# 将时间坐标添加到数据集中
combined_ds = combined_ds.assign_coords(time=('time', time_coords))
combined_ds = combined_ds.rename({'Band1': 'vqi'})
#combined_ds["GPP"] = combined_ds["GPP"] * 0.001 * 0.001
# 如果你想保存合并后的数据到新的NC文件中
output_nc = r'E:\530\VQI\vqi\combined_data.nc'
combined_ds.to_netcdf(output_nc)
print(f"Combined dataset saved to {output_nc}")
def combin_tif_to_nc():
'''
第一步，先进行tif到nc的转换
:return:
'''
# 设置输入和输出文件路径
input_dir = r'E:\530\VQI' # 包含TIFF文件的目录
input_list = Search_File(input_dir,".tif")
for p,n in input_list:
# 使用gdal.Warp()函数进行转换，这儿可以设定转换出的文件行列号
gdal.Warp(r"E:\530\VQI\vqi\{}.nc".format(n),p,height=553,width=1638,format='NetCDF')
print("转换完成！")
def mk_test():
import xarray as xr
import numpy as np
import pymannkendall as mk
# 打开NetCDF文件
ds = xr.open_dataset(r'E:\530\VQI\vqi\combined_data.nc')
# 获取数据和坐标
data = ds['vqi'].values
times = ds['time'].values
lats = ds['lat'].values
lons = ds['lon'].values
# 创建空的结果数组
sen_value = np.zeros_like(data[0]) * np.nan # 初始化为NaN
trend_value = np.zeros_like(data[0]) * np.nan
h_value = np.zeros_like(data[0]) * np.nan
# 遍历每个网格点
for i in range(data.shape[1]):
for j in range(data.shape[2]):
# 获取当前网格点的时间序列
ts = data[:, i, j]
print(ts)
# 检查是否有缺失值
if np.any(ts == -3.4e+38):
continue # 跳过当前网格点
if np.isnan(ts).any():
continue
# 进行MK检验
trend, h, p, z, Tau, s, var_s, slope, intercept = mk.original_test(ts)
if trend == "decreasing":
trend_id = -1
elif trend == "increasing":
trend_id = 1
elif trend == "no trend":
trend_id = 0
# 存储结果
sen_value[i, j] = slope
trend_value[i, j] = trend_id
h_value[i, j] = h
# 创建NetCDF文件
ds_out = xr.Dataset({'mk_trend': (['y', 'x'], trend_value),
'mk_sen' : (['y', 'x'], sen_value),
"mk_h" : (['y', 'x'], h_value)},
coords={'y': lats, 'x': lons})
ds_out.to_netcdf(r'E:\530\VQI\vqi\mk_results.nc')
print('Successfully created mk_results.nc')
# first
# combin_tif_to_nc()
# second
# combin()
# end
mk_test()