03_frequency-matching.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "d7c3822a-3fb4-4e1c-9cf4-c60f57bbb147",
   "metadata": {},
   "source": [
    "# 频率匹配方法（Frequency-matching method）\n",
    "\n",
    "#### —— 数值模式降水系统性偏差订正技术\n",
    "\n",
    "国家气象中心天气预报技术研发室  \n",
    "代刊, 王玉, 唐健, June, 2021  \n",
    "\n",
    "\n",
    "#### Objectives\n",
    "* 了解频率匹配方法的原理；\n",
    "* 掌握频率匹配方法中观测和预报累积概率密度曲线CDF如何构建；\n",
    "* 掌握如何利用CDF进行降水偏差订正；\n",
    "* 了解频率匹配方法的业务实际应用；\n",
    "* 了解分位值映射方法；\n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e1b6ecb8-3c00-4eef-90b4-0c919c5c4c47",
   "metadata": {},
   "outputs": [],
   "source": [
    "# set up things\n",
    "%matplotlib inline\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "3656786a-88f2-4a88-9ec8-80bbf3789f05",
   "metadata": {},
   "outputs": [],
   "source": [
    "# load necessary libraries\n",
    "# you should install cartopy with 'conda install -c conda-forge cartopy'\n",
    "import sys\n",
    "import pickle\n",
    "from pathlib import Path\n",
    "import xarray as xr\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import cartopy.crs as ccrs\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from nmc_met_base.qpf_cal import cum_freq, cfd_match"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2b24047d-78a5-49ed-90c5-6ee104fb756f",
   "metadata": {},
   "source": [
    "##  频率匹配方法原理\n",
    "---\n",
    "\n",
    "由于降水具有非高斯分布特征，其系统性误差并不能通过观测与预报之间的差值直接订正。为此，朱跃建(2015)发展了频率匹配方法，通过改变降水预报量值，使得预报的降水频率分布向观测频率靠拢。具体步骤分为两步：\n",
    "\n",
    "* 第一步，需要采用训练数据集构建不同百分位预报值$F$的累积频率分布曲线$FCDF$和相应的观测累积频率分布$OCDF$。\n",
    "* 第二步，预报调整，即对预报值进行替换，使得CDF曲线上预报与观测的频率值的相同。\n",
    "\n",
    "<img src=\"./images/freq_match_01.png\" alt=\"image\" style=\"height: 350px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d65c69d2-3825-478c-8259-ecbeaf403603",
   "metadata": {},
   "source": [
    "## 构建累积频率分布曲线\n",
    "---\n",
    "\n",
    "* 累积频率分布曲线的构建方法为，预先给定一系列升序排列的降水阈值，然后针对每个阈值计算某区域内，超过该阈值的预报或观测样本数所占总样本数据的比例，得到该阈值的累积频率值CDF，连接阈值序列的CDF即可得到累积频率分布曲线CDFs。\n",
    "* CDFs的形状与气候背景关系密切，为反映CDFs的区域性特点，但同时保证降水事件的数量（尤其是大量级降水事件）不会太低。"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6791476d-ea27-4e86-99d8-ff776a995c0e",
   "metadata": {},
   "source": [
    "### 对比站点的观测与预报不同等级降水的发生频率\n",
    "\n",
    "* `./samples/station_obs_fcst/`目录下放置了北京、上海、广州、成都站及其相似站点的预报和观测实况数据；\n",
    "* 数据存放方式为：一个点/起报时刻/预报时效一个文件，时间段为2019年6月1日至8月31日；\n",
    "* 读入统一个预报时效的所有数据，构建样本集放入Pandas表格\n",
    "* 下面以北京站为例，对比观测和预报的不同等级降水发生频率"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "d9672e48-12b7-4c30-bb3b-3ce310caad2b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#\n",
    "# retrieve observation and forecast data\n",
    "datafiles = Path(\"./samples/station_obs_fcst/beijing/\").glob(\"*_024.txt\")\n",
    "data = pd.DataFrame()\n",
    "for datafile in datafiles:\n",
    "    tmp = pd.read_csv(datafile)\n",
    "    data = pd.concat([data, tmp], axis=0, ignore_index=True)\n",
    "data.columns = [s.strip() for s in data.columns]\n",
    "data['OBS'] = data['OBS'].astype(float)\n",
    "data['FCST'] = data['FCST'].astype(float)\n",
    "\n",
    "# compare the frequency of obs and forecast\n",
    "bins = [0,0.1,10,25,50,100,250]\n",
    "obs_hist, _ = np.histogram(data['OBS'], bins=bins)\n",
    "fct_hist, _ = np.histogram(data['FCST'], bins=bins)\n",
    "\n",
    "plt.figure(figsize=(10,6))\n",
    "plt.rcParams.update({\"font.size\":20})\n",
    "plt.bar(np.arange(len(bins)-1)-0.2, obs_hist, 0.4, label=\"Observation\")\n",
    "plt.bar(np.arange(len(bins)-1)+0.2, fct_hist, 0.4, label=\"Forecast\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "8ee51932-5859-4ea2-bf39-cdd35884cf84",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#\n",
    "# retrieve observation and forecast data\n",
    "datafiles = Path(\"./samples/station_obs_fcst/beijing/\").glob(\"*_144.txt\")\n",
    "data = pd.DataFrame()\n",
    "for datafile in datafiles:\n",
    "    tmp = pd.read_csv(datafile)\n",
    "    data = pd.concat([data, tmp], axis=0, ignore_index=True)\n",
    "data.columns = [s.strip() for s in data.columns]\n",
    "data['OBS'] = data['OBS'].astype(float)\n",
    "data['FCST'] = data['FCST'].astype(float)\n",
    "\n",
    "# compare the frequency of obs and forecast\n",
    "bins = [0,0.1,10,25,50,100,250]\n",
    "obs_hist, _ = np.histogram(data['OBS'], bins=bins)\n",
    "fct_hist, _ = np.histogram(data['FCST'], bins=bins)\n",
    "\n",
    "plt.figure(figsize=(10,6))\n",
    "plt.rcParams.update({\"font.size\":20})\n",
    "plt.bar(np.arange(len(bins)-1)-0.2, obs_hist, 0.4, label=\"Observation\")\n",
    "plt.bar(np.arange(len(bins)-1)+0.2, fct_hist, 0.4, label=\"Forecast\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b386157f-488c-4e79-a94b-bc921f1e3fa1",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\">\n",
    "    <b>EXERCISE</b>:\n",
    "    对比分析不同站点、不同时效的降水出现频率:\n",
    "     <ul>\n",
    "      <li>对比不同站点、不同等级的观测、预报降水出现频率.</li>\n",
    "      <li>对比24h、96h、168h、240h时效，不同等级的观测、预报降水出现频率.</li>\n",
    "    </ul>\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f75818c4-5c7d-43f8-b620-bf8e071590b6",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-info\">\n",
    "    <b>SOLUTION</b>\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "35f0e2ec-3f6b-4aef-be24-20ad5c87e0da",
   "metadata": {},
   "source": [
    "### 构建累积概率分布函数"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9cb3e5f3-e437-490c-a6b2-1e75c12c7c55",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\">\n",
    "    <b>EXERCISE</b>:\n",
    "    累积概率分布函数的构建和分析:\n",
    "     <ul>\n",
    "      <li>构建累积概率分布函数程序\"cum_freq\".</li>\n",
    "      <li>对比分析不同站点、不同时效的观测和预报累积概率分布曲线.</li>\n",
    "    </ul>\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0ba3429-2ee1-42d2-96c2-f0e29b225118",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-info\">\n",
    "    <b>SOLUTION</b>\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "48a20cf8-c1f1-4548-a48c-c8d5bce3bac6",
   "metadata": {},
   "source": [
    "```python\n",
    "def cum_freq(data, bins=None, norm=True):\n",
    "    \"\"\"Calculate the cumulative frequency distribution.\n",
    "\n",
    "    Arguments:\n",
    "        data {numpy nd-array} -- numpy nd-array. The missing value is allowed.\n",
    "\n",
    "    Keyword Arguments:\n",
    "        bins {numpy array} -- the bin-edegs used to calculate CFD.\n",
    "        norm {bool} -- normalize the distribution (default: {True})\n",
    "    \"\"\"\n",
    "\n",
    "    # set the bin edges\n",
    "    if bins is None:\n",
    "        bins = np.concatenate(([0.1, 1], np.arange(2, 10, 1),\n",
    "                               np.arange(10, 152, 2)))\n",
    "\n",
    "    # TO Do, complete the code\n",
    "\n",
    "    # return the bin edges and CFD\n",
    "    return cfd_array, bins\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "4a0cc4d3-afe6-47fa-834a-0c0abcf2a6de",
   "metadata": {},
   "outputs": [],
   "source": [
    "#\n",
    "# retrieve observation and forecast data\n",
    "datafiles = Path(\"./samples/station_obs_fcst/guangzhou/\").glob(\"*_048.txt\")\n",
    "data = pd.DataFrame()\n",
    "for datafile in datafiles:\n",
    "    tmp = pd.read_csv(datafile)\n",
    "    data = pd.concat([data, tmp], axis=0, ignore_index=True)\n",
    "data.columns = [s.strip() for s in data.columns]\n",
    "data['OBS'] = data['OBS'].astype(float)\n",
    "data['FCST'] = data['FCST'].astype(float)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "c069a88e-76f9-4bf0-9a8a-c99c22c804b1",
   "metadata": {},
   "outputs": [],
   "source": [
    "#\n",
    "# calculate the CFD\n",
    "obs_cfd, bins = cum_freq(data['OBS'])\n",
    "fct_cfd, bins = cum_freq(data['FCST'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "719def2a-364f-44e7-9f71-8ceb594c74c5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#\n",
    "# Compare the Observation and Forecasting CFD\n",
    "plt.figure(figsize=(10,6))\n",
    "plt.rcParams.update({\"font.size\":20, \"axes.linewidth\":3, \"xtick.major.width\":3, \"ytick.major.width\":3})\n",
    "plt.plot(bins, obs_cfd, label=\"Observation\", linewidth=2)\n",
    "plt.plot(bins, fct_cfd, label=\"Forecast\", linewidth=2)\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4bb1dcf8-a11f-4f69-80b2-c498f7b74921",
   "metadata": {},
   "source": [
    "## 基于CFD曲线进行预报调整\n",
    "---\n",
    "\n",
    "* 对于一个原始预报值“RAW”可以通过预报频率分布曲线（实线）获得出现频率值，为保证预报与观测具有相同频率分布，则在观测频率分布曲线（虚线）上找到同样频率的位置，其对应的观测量级即为订正后的预报值“CAL”。通过此方法，可以看到在“RAW”值上预报频率高于观测（过度预报），通过频率匹配订正可以适当减小预报量级。"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "177f3ba4-208f-4633-b5c2-dfcc5a17b0ca",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\">\n",
    "    <b>EXERCISE</b>:\n",
    "    进行预报调整:\n",
    "     <ul>\n",
    "      <li>构建累积概率分布函数匹配调整程序\"cfd_match\".</li>\n",
    "      <li>对某个城市、某个时效的预报量进行CFD匹配调整，使得观测与预报的CFD一致.</li>\n",
    "    </ul>\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b1288766-4b77-4bf5-bb96-f28b4a0f2363",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-info\">\n",
    "    <b>SOLUTION</b>\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "85b53405-eaa9-4bf2-9702-c43525444c7e",
   "metadata": {},
   "source": [
    "```python\n",
    "def cfd_match(data, cfd_obs, cfd_fcst, bins):\n",
    "    \"\"\"\n",
    "    Perform the frequency-matching methods.\n",
    "\n",
    "    Args:\n",
    "        data (np.array): forecast data to be calibrated.\n",
    "        cdf_obs (np.array): 1D array, the cumulative frequency distribution of observations.\n",
    "        cdf_fcst (np.array): 1D array, the cumulative frequency distribution of forecasts.\n",
    "        bins (np.array): 1D array, bin edges for CFD.\n",
    "\n",
    "    Returns:\n",
    "        np.array: calibrated forecasts.\n",
    "    \"\"\"\n",
    "\n",
    "    # To DO, complete the code\n",
    "    \n",
    "    return data_cal\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "621f24b6-e3bb-48b2-99af-ae0daa31d6bf",
   "metadata": {},
   "outputs": [],
   "source": [
    "#\n",
    "# retrieve observation and forecast data\n",
    "datafiles = Path(\"./samples/station_obs_fcst/guangzhou/\").glob(\"*_048.txt\")\n",
    "data = pd.DataFrame()\n",
    "for datafile in datafiles:\n",
    "    tmp = pd.read_csv(datafile)\n",
    "    data = pd.concat([data, tmp], axis=0, ignore_index=True)\n",
    "data.columns = [s.strip() for s in data.columns]\n",
    "data['OBS'] = data['OBS'].astype(float)\n",
    "data['FCST'] = data['FCST'].astype(float)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fbefbf83-bf9e-47f4-831b-d2a6cbfb9568",
   "metadata": {},
   "source": [
    "假定预报值为`[0, 0.1, 2, 6, 15, 52, 100, 130, 203.]`，计算订正后的降水值"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "6f9acf80-ecc2-478f-abed-d84929a2794c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[  0.      0.1     0.1     1.305  11.978  65.2   104.    130.667 203.   ]\n"
     ]
    }
   ],
   "source": [
    "with np.printoptions(precision=3, suppress=True):\n",
    "    print(cfd_match([0, 0.1, 2, 6, 15, 52, 100, 130, 203.], obs_cfd, fct_cfd, bins))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "80794476-a4be-417b-a98f-d1940db34e45",
   "metadata": {},
   "source": [
    "对预报值进行频率匹配调整，并对比调整后的预报和观测的CFD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "da02eb03-322a-490e-89b0-4b261d20af13",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#\n",
    "# retrieve observation and forecast data\n",
    "datafiles = Path(\"./samples/station_obs_fcst/guangzhou/\").glob(\"*_048.txt\")\n",
    "data = pd.DataFrame()\n",
    "for datafile in datafiles:\n",
    "    tmp = pd.read_csv(datafile)\n",
    "    data = pd.concat([data, tmp], axis=0, ignore_index=True)\n",
    "data.columns = [s.strip() for s in data.columns]\n",
    "data['OBS'] = data['OBS'].astype(float)\n",
    "data['FCST'] = data['FCST'].astype(float)\n",
    "\n",
    "# adjust the forecast values\n",
    "fct_cal = cfd_match(data['FCST'], obs_cfd, fct_cfd, bins)\n",
    "\n",
    "# calculate the calibrated forecast and observation CFD\n",
    "obs_cfd, bins = cum_freq(data['OBS'])\n",
    "fct_cal_cfd, bins = cum_freq(fct_cal)\n",
    "\n",
    "# Compare\n",
    "plt.figure(figsize=(10,6))\n",
    "plt.rcParams.update({\"font.size\":20, \"axes.linewidth\":3, \"xtick.major.width\":3, \"ytick.major.width\":3})\n",
    "plt.plot(bins, obs_cfd, label=\"Observation\", linewidth=2)\n",
    "plt.plot(bins, fct_cal_cfd, label=\"Calibrated Forecast\", linewidth=2)\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c1a8a19a-e563-4fff-a4c5-43fae9e71012",
   "metadata": {},
   "source": [
    "## 分位值映射(Quantile Mapping)\n",
    "---\n",
    "\n",
    "分位值映射是另一种较为常用的降水量级订正方法，不仅用于天气模式（Bremnes，2007），也多用于气候模式（Maraun，2013）。分位值映射具有与频率匹配方法相同的两个步骤，主要不同点在于其采用降水样本的分位值分布曲线。Maraun（2013）指出当观测数据的分辨率远高于预报数据时，分位值映射还起到将低分辨预报转换为高分辨率的功能，但该降尺度作用会使得校正后降水序列的时空结构歪曲（如区域平均的极端值被过高估计，降水变化趋势受到影响等）；为避免这些问题，需要加入随机偏差订正。Cannon等（2015）进一步引入分位差值映射来确保订正后降水时间序列的变化趋势不变，并将单变量的分位值映射拓展到多变量（降水和温度）协同订正。Hamill等（2017）以分位值映射方法为基础，发展多中心集合模式的降水概率预报后处理技术，得到的结果表现出更高的预报技巧和可靠性，同时保留了解析度和空间分布细节。\n",
    "\n",
    "<img src=\"./images/quantile_mapping.png\" alt=\"image\" style=\"height: 600px;\">  \n",
    "<div id=\"images\">\n",
    "<img src=\"./images/qmap_raw.png\" alt=\"image\" style=\"height: 300px;\">  \n",
    "<img src=\"./images/qmap_cal.png\" alt=\"image\" style=\"height: 300px;\"> \n",
    "</div> "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4063d9bf-56eb-430a-a4e2-777d4c7a52ef",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "620027cf-844c-4b4e-8631-9101806dc887",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\">\n",
    "    <b>EXERCISE</b>:\n",
    "    了解分位值映射订正技术:\n",
    "     <ul>\n",
    "      <li>掌握从Python中调用R语言函数库实现分位值映射.</li>\n",
    "      <li>对某个城市、某个时效的预报量进行分位值映射调整，并对比观测与预报的CFD.</li>\n",
    "    </ul>\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fbe9accc-3fa6-42b7-b37a-4742570a326f",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-info\">\n",
    "    <b>SOLUTION</b>\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "88ded8bf-2c07-422a-b61f-63563c1f1117",
   "metadata": {},
   "outputs": [],
   "source": [
    "# load rpy2 functions\n",
    "import rpy2.robjects as ro\n",
    "from rpy2.robjects.packages import importr\n",
    "from rpy2.robjects import pandas2ri\n",
    "from rpy2.robjects.conversion import localconverter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "aaa5d04b-856b-461b-9566-fd335209dcba",
   "metadata": {},
   "outputs": [],
   "source": [
    "# import R's \"qmap\" package\n",
    "qmap = importr(\"qmap\")\n",
    "\n",
    "# load examples data\n",
    "ro.r(\"data(obsprecip)\")\n",
    "ro.r(\"data(modprecip)\")\n",
    "with localconverter(ro.default_converter + pandas2ri.converter):\n",
    "    obsprecip = ro.conversion.rpy2py(ro.r['obsprecip'])\n",
    "    modprecip = ro.conversion.rpy2py(ro.r['modprecip'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "9a6f827b-f898-4af7-8aa8-1ae05ecd8948",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>MOSS</th>\n",
       "      <th>GEIRANGER</th>\n",
       "      <th>BARKESTAD</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1-1-1961</th>\n",
       "      <td>0.1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2-1-1961</th>\n",
       "      <td>0.2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3-1-1961</th>\n",
       "      <td>0.9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4-1-1961</th>\n",
       "      <td>10.6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5-1-1961</th>\n",
       "      <td>1.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27-12-1990</th>\n",
       "      <td>23.0</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28-12-1990</th>\n",
       "      <td>4.0</td>\n",
       "      <td>4.4</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29-12-1990</th>\n",
       "      <td>24.5</td>\n",
       "      <td>1.2</td>\n",
       "      <td>0.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30-12-1990</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2.5</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31-12-1990</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.7</td>\n",
       "      <td>0.8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10957 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            MOSS  GEIRANGER  BARKESTAD\n",
       "1-1-1961     0.1        0.0        0.0\n",
       "2-1-1961     0.2        0.0        0.0\n",
       "3-1-1961     0.9        0.0        0.0\n",
       "4-1-1961    10.6        0.0        0.0\n",
       "5-1-1961     1.5        0.0        0.0\n",
       "...          ...        ...        ...\n",
       "27-12-1990  23.0        0.1        0.0\n",
       "28-12-1990   4.0        4.4        0.0\n",
       "29-12-1990  24.5        1.2        0.1\n",
       "30-12-1990   0.0        2.5        0.0\n",
       "31-12-1990   0.0        0.7        0.8\n",
       "\n",
       "[10957 rows x 3 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "obsprecip "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "9e60345a-8f08-439b-9ae3-6357037f0319",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>MOSS</th>\n",
       "      <th>GEIRANGER</th>\n",
       "      <th>BARKESTAD</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2-1-1961</th>\n",
       "      <td>2.28300</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>3.17700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3-1-1961</th>\n",
       "      <td>2.44300</td>\n",
       "      <td>10.8600</td>\n",
       "      <td>1.71900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4-1-1961</th>\n",
       "      <td>3.09900</td>\n",
       "      <td>12.7300</td>\n",
       "      <td>6.63600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5-1-1961</th>\n",
       "      <td>0.00000</td>\n",
       "      <td>9.7720</td>\n",
       "      <td>9.67600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6-1-1961</th>\n",
       "      <td>0.14000</td>\n",
       "      <td>0.6448</td>\n",
       "      <td>7.11000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26-12-1990</th>\n",
       "      <td>3.91200</td>\n",
       "      <td>7.2530</td>\n",
       "      <td>0.51610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27-12-1990</th>\n",
       "      <td>0.04119</td>\n",
       "      <td>19.8300</td>\n",
       "      <td>6.20400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28-12-1990</th>\n",
       "      <td>1.97700</td>\n",
       "      <td>17.0200</td>\n",
       "      <td>10.13000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29-12-1990</th>\n",
       "      <td>1.09600</td>\n",
       "      <td>16.3300</td>\n",
       "      <td>3.55200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30-12-1990</th>\n",
       "      <td>0.01888</td>\n",
       "      <td>4.8650</td>\n",
       "      <td>0.02652</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10799 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "               MOSS  GEIRANGER  BARKESTAD\n",
       "2-1-1961    2.28300     0.0000    3.17700\n",
       "3-1-1961    2.44300    10.8600    1.71900\n",
       "4-1-1961    3.09900    12.7300    6.63600\n",
       "5-1-1961    0.00000     9.7720    9.67600\n",
       "6-1-1961    0.14000     0.6448    7.11000\n",
       "...             ...        ...        ...\n",
       "26-12-1990  3.91200     7.2530    0.51610\n",
       "27-12-1990  0.04119    19.8300    6.20400\n",
       "28-12-1990  1.97700    17.0200   10.13000\n",
       "29-12-1990  1.09600    16.3300    3.55200\n",
       "30-12-1990  0.01888     4.8650    0.02652\n",
       "\n",
       "[10799 rows x 3 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "modprecip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "7aafc2ff-9f12-4b4d-afd2-9526659b72c4",
   "metadata": {},
   "outputs": [],
   "source": [
    "with localconverter(ro.default_converter + pandas2ri.converter):\n",
    "    qm_fit = qmap.fitQmapQUANT(obsprecip['GEIRANGER'], modprecip['GEIRANGER'], qstep=0.1, nboot=1, wet_day=True)\n",
    "    qm = qmap.doQmapQUANT(modprecip['GEIRANGER'],qm_fit,type=\"linear\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "dfb6f28f-0e03-45f7-84c6-2c0c27ce2824",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "$par\n",
      "$par$modq\n",
      "           [,1]\n",
      " [1,]  1.838798\n",
      " [2,]  2.708021\n",
      " [3,]  3.679105\n",
      " [4,]  4.833817\n",
      " [5,]  6.217873\n",
      " [6,]  7.952622\n",
      " [7,] 10.160000\n",
      " [8,] 12.990000\n",
      " [9,] 16.690527\n",
      "[10,] 23.308633\n",
      "[11,] 90.300000\n",
      "\n",
      "$par$fitq\n",
      "             [,1]\n",
      " [1,]  0.09297401\n",
      " [2,]  0.20000000\n",
      " [3,]  0.60000000\n",
      " [4,]  1.20000000\n",
      " [5,]  2.00000000\n",
      " [6,]  3.10000000\n",
      " [7,]  4.60000000\n",
      " [8,]  6.70000000\n",
      " [9,] 10.50000000\n",
      "[10,] 17.89147579\n",
      "[11,] 75.40000000\n",
      "\n",
      "\n",
      "$wet.day\n",
      "[1] 1.838798\n",
      "\n",
      "attr(,\"class\")\n",
      "[1] \"fitQmapQUANT\"\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(qm_fit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "28523ba9-78f1-44a4-b92c-4afcea8f213b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.          5.11943463  6.50706714 ... 10.86797385 10.12978152\n",
      "  1.21802387]\n"
     ]
    }
   ],
   "source": [
    "print(qm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "3e815eb7-8f8e-47eb-bf28-84a872aa1a39",
   "metadata": {},
   "outputs": [],
   "source": [
    "#\n",
    "# 定义计算分位值的函数\n",
    "def sqrtqunat(x, qstep=0.01):\n",
    "    qq = np.quantile(x, np.arange(0,1,qstep))\n",
    "    return np.sqrt(qq)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "bdb5b674-088b-4d2e-abad-dfa35bcde67c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,10))\n",
    "plt.rcParams.update({\"font.size\":20, \"axes.linewidth\":3, \"xtick.major.width\":3, \"ytick.major.width\":3})\n",
    "plt.plot(sqrtqunat(modprecip['GEIRANGER']), sqrtqunat(obsprecip['GEIRANGER']), label=\"Raw\", marker='o', linewidth=2)\n",
    "plt.plot(sqrtqunat(qm), sqrtqunat(obsprecip['GEIRANGER']), label=\"Calibrated\", marker='o', linewidth=2)\n",
    "plt.xlim([0,7])\n",
    "plt.ylim([0,7])\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "39fc88cc-acaa-4db4-ae43-18853a8b6528",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from nmc_met_base.qpf_cal import cum_freq\n",
    "\n",
    "#\n",
    "# calculate the CFD\n",
    "obs_cfd, bins = cum_freq(obsprecip['GEIRANGER'])\n",
    "fct_cfd, bins = cum_freq(modprecip['GEIRANGER'])\n",
    "fct_cal_cfd, bins = cum_freq(qm)\n",
    "\n",
    "#\n",
    "# Compare the Observation and Forecasting CFD\n",
    "plt.figure(figsize=(10,6))\n",
    "plt.rcParams.update({\"font.size\":20, \"axes.linewidth\":3, \"xtick.major.width\":3, \"ytick.major.width\":3})\n",
    "plt.plot(bins, obs_cfd, label=\"Observation\", linewidth=2)\n",
    "plt.plot(bins, fct_cfd, label=\"Forecast\", linewidth=2)\n",
    "plt.plot(bins, fct_cal_cfd, label=\"Calibrated Forecast\", linewidth=2)\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "051e4811-4367-4af3-a1cd-870768a9b562",
   "metadata": {},
   "source": [
    "## 频率匹配订正技术的业务应用\n",
    "---\n",
    "\n",
    "<img src=\"./images/freq_match_02.png\" alt=\"image\" style=\"height: 200px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1658243e-8707-4def-a0a3-d5cb173009f0",
   "metadata": {},
   "source": [
    "## References\n",
    "\n",
    "* Zhu, Y. J. and Y. Luo (2015). \"Precipitation Calibration Based on the Frequency-Matching Method.\" weather and forecasting 30(5): 1109-1124."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ad75121a-bd70-4967-91a9-ee3844884f0a",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Meteorology with Python",
   "language": "python",
   "name": "met"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.10"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {},
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}