使用cfgrib将数据保存为GRIB文件

import cfgrib as cf
from pprint import pprint

ds = cf.open_datasets("surface.grib")[-1]
pprint(ds.attrs)
pprint(ds["skt"].attrs)

import cfgrib as cf
from cfgrib.xarray_to_grib import to_grib
import numpy as np
from xarray import DataArray, Dataset
from pandas import to_datetime


if __name__ == '__main__':
    # 生成假数据
    data = np.random.randint(280, 303, (2, 10, 10))

    # 定义一个DataArray变量，并顺便定义GRIB参数
    data_array = DataArray(
        data=data, dims=["time", "latitude", "longitude"],
        coords={
            "time": to_datetime(["2020-01-01 00:00", "2020-01-01 01:00"]),
            "latitude": np.arange(10, 0, -1),
            "longitude": np.arange(0, 100, 10)
        }, attrs={
            "units": "K",
            "long_name": "Sea surface temperature",
            "standard_name": "Sea surface temperature",
            # The following keys and values will be used in GRIB.
            "GRIB_paramId": 34,		# sst在ECMWF的GRIB码表中的ID是34
            "GRIB_shortName": "sst",	# shortName应确保是正确对应的
            "GRIB_units": "K",	# units应确保是正确对应的
            "GRIB_name": "Sea surface temperature",		# name应确保是正确对应的
            "GRIB_stepUnits": 1,
            "GRIB_stepType": "instant",
            "GRIB_gridType": "regular_ll",	# 投影方式，regular_ll代表等经纬度投影
            "GRIB_iDirectionIncrementInDegrees": 10,	# 经度方向格点间隔，这个值的正负和下面的iScanNegatively相关。
            "GRIB_iScanNegatively": 0,
            "GRIB_jDirectionIncrementInDegrees": 1,		# 纬度方向格点间隔，这个值的正负和下面的jScanPositively相关。
            "GRIB_jScanPositively": 0,
            "GRIB_latitudeOfFirstGridPointInDegrees": 10,	# 区域的经纬度起始点数值
            "GRIB_latitudeOfLastGridPointInDegrees": 1,		# 区域的经纬度起始点数值
            "GRIB_longitudeOfFirstGridPointInDegrees": 0,	# 区域的经纬度起始点数值
            "GRIB_longitudeOfLastGridPointInDegrees": 90,	# 区域的经纬度起始点数值
            "GRIB_Ny": 10,		# 纬度方向的格点数
            "GRIB_Nx": 10,		# 经度方向的格点数
            "GRIB_typeOfLevel": "surface",		# 变量的level类型
            # The following keys and values can't be found at ECMWF websites.
            "GRIB_cfName": "unknown",
            "GRIB_cfVarName": "sst",
            "GRIB_dataType": "an",  # Analysis data, defined at https://codes.ecmwf.int/grib/format/mars/type/
            "GRIB_gridDefinitionDescription": "Latitude/Longitude Grid",
            "GRIB_numberOfPoints": 100,		# 每一时刻数据场的总格点数，也就是 Nx * Ny
            "GRIB_totalNumber": 0,
            "GRIB_uvRelativeToGird": 0
        }
    )

    dataset = Dataset({
        "sst": data_array
    }, attrs={
        "GRIB_centre": "ecmf",		# 定义使用哪个机构的码表
        "GRIB_edition": 1			# 定义GRIB版本号
    })

    to_grib(dataset, "data/sst_test.grib")

    ds = cf.open_dataset("data/sst_test.grib")
    sst = ds["sst"]

    print(np.where(data_array.to_numpy() != sst.to_numpy()))
	print("")
    pprint(ds.attrs)
	print("")
    pprint(sst.attrs)

dataset = Dataset({
    "sst": data_array
}, attrs={
    "GRIB_centre": "my_centre",		# 换成一个自定义的centre名称
})
    
dataset = Dataset({
    "sst": data_array		# 干脆不设置attrs
})