.. DO NOT EDIT. .. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY. .. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE: .. "auto_examples/qpf_manila.py" .. LINE NUMBERS ARE GIVEN BELOW. .. only:: html .. note:: :class: sphx-glr-download-link-note Click :ref:`here ` to download the full example code .. rst-class:: sphx-glr-example-title .. _sphx_glr_auto_examples_qpf_manila.py: QPF (Manila) ======================================================== This example demonstrates how to perform operational deterministic QPF up to three hours using raingauge data from Manila and radar data from Subic. .. GENERATED FROM PYTHON SOURCE LINES 10-13 Definitions -------------------------------------------------------- .. GENERATED FROM PYTHON SOURCE LINES 13-41 .. code-block:: default import os import numpy as np import pandas as pd from pyresample import utils import xarray import cartopy.feature as cfeature import cartopy.crs as ccrs import matplotlib import imageio import matplotlib.pyplot as plt from matplotlib.colors import BoundaryNorm from swirlspy.rad.uf_ph import read_uf_ph from swirlspy.qpe.utils import dbz2rr, rr2rf, locate_file, timestamps_ending from swirlspy.qpe.utils import multiple_acc from swirlspy.obs.rain import Rain from swirlspy.qpf.rover import rover from swirlspy.qpf.sla import sla from swirlspy.core.resample import grid_resample plt.switch_backend('agg') THIS_DIR = os.getcwd() os.chdir(THIS_DIR) start_time = pd.Timestamp.now() .. rst-class:: sphx-glr-script-out .. code-block:: pytb Traceback (most recent call last): File "/tmp/build/docs/swirlspy/swirlspy/examples/qpf_manila.py", line 30, in from swirlspy.qpf.rover import rover File "/tmp/build/docs/swirlspy/swirlspy/qpf/rover.py", line 6, in from rover.rover import rover as rover_api ImportError: libopencv_core.so.3.4: cannot open shared object file: No such file or directory .. GENERATED FROM PYTHON SOURCE LINES 42-50 Initialising --------------------------------------------------- This section demonstrates extracting radar reflectivity data. Step 1: Define a basetime. .. GENERATED FROM PYTHON SOURCE LINES 50-54 .. code-block:: default # Supply basetime basetime = pd.Timestamp('20180811112000').floor('min') .. GENERATED FROM PYTHON SOURCE LINES 55-57 Step 2: Using basetime, generate timestamps of desired radar files timestamps_ending() and locate files using locate_file(). .. GENERATED FROM PYTHON SOURCE LINES 57-70 .. code-block:: default # Obtain radar files dir = THIS_DIR + '/../tests/samples/uf_ph/sub/' located_files = [] radar_ts = timestamps_ending( basetime, duration=pd.Timedelta(minutes=60), interval=pd.Timedelta(minutes=10) ) for timestamp in radar_ts: located_files.append(locate_file(dir, timestamp)) .. GENERATED FROM PYTHON SOURCE LINES 71-72 Step 3: Define the target grid as a pyresample AreaDefinition. .. GENERATED FROM PYTHON SOURCE LINES 72-90 .. code-block:: default area_id = "epsg3123_240km" description = ("A 240 m resolution rectangular grid " "centred at Subic RADAR and extending to 240 km " "in each direction") proj_id = 'epsg3123' projection = ('+proj=tmerc +lat_0=0 ' '+lon_0=121 +k=0.99995 +x_0=500000 ' '+y_0=0 +ellps=clrk66 +towgs84=-127.62,-67.24,' '-47.04,-3.068,4.903,1.578,-1.06 +units=m ' '+no_defs') x_size = 500 y_size = 500 area_extent = (191376.04113, 1399386.68659, 671376.04113, 1879386.68659) area_def = utils.get_area_def( area_id, description, proj_id, projection, x_size, y_size, area_extent ) .. GENERATED FROM PYTHON SOURCE LINES 91-94 Step 4: Read data from radar files into xarray.DataArray using read_uf_ph(). .. GENERATED FROM PYTHON SOURCE LINES 94-104 .. code-block:: default reflectivity_list = [] # stores reflec from read_iris() for filename in located_files: reflec = read_uf_ph( filename, area_def=area_def, coord_label=['easting', 'northing'], indicator='cappi', elevation=2 ) reflectivity_list.append(reflec) .. GENERATED FROM PYTHON SOURCE LINES 105-107 Step 5: Assigning reflectivity xarrays at the last two timestamps to variables for use during ROVER QPF. .. GENERATED FROM PYTHON SOURCE LINES 107-113 .. code-block:: default xarray1 = reflectivity_list[-2] xarray2 = reflectivity_list[-1] initialising_time = pd.Timestamp.now() .. GENERATED FROM PYTHON SOURCE LINES 114-120 Running ROVER and Semi-Lagrangian Advection ------------------------------------------- 1. Concatenate two reflectivity xarrays along time dimension. 2. Run ROVER, with the concatenated xarray as the input. 3. Perform Semi-Lagrangian Advection using the motion fields from rover. .. GENERATED FROM PYTHON SOURCE LINES 120-140 .. code-block:: default # Combining the two reflectivity DataArrays # the order of the coordinate keys is now ['y', 'x', 'time'] # as opposed to ['time', 'x', 'y'] reflec_concat = xarray.concat([xarray1, xarray2], dim='time') # Rover motion_u, motion_v = rover( reflec_concat ) rover_time = pd.Timestamp.now() # Semi Lagrangian Advection reflectivity = sla( reflec_concat, motion_u, motion_v, steps=30 ) sla_time = pd.Timestamp.now() .. GENERATED FROM PYTHON SOURCE LINES 141-147 Concatenating observed and forecasted reflectivities ------------------------------------------------------- 1. Add forecasted reflectivity to reflectivity_list. 2. Concatenate observed and forecasted reflectivity xarray.DataArrays along the time dimension. .. GENERATED FROM PYTHON SOURCE LINES 147-153 .. code-block:: default reflectivity_list.append(reflectivity[1:, ...]) reflectivity = xarray.concat(reflectivity_list, dim='time') concat_time = pd.Timestamp.now() .. GENERATED FROM PYTHON SOURCE LINES 154-162 Generating radar reflectivity maps ----------------------------------- Define the color scale and format of the plots and plot using xarray.plot(). In this example, only hourly images will be plotted. .. GENERATED FROM PYTHON SOURCE LINES 162-241 .. code-block:: default # Defining colour scale and format levels = [ -32768, 10, 15, 20, 24, 28, 32, 34, 38, 41, 44, 47, 50, 53, 56, 58, 60, 62 ] cmap = matplotlib.colors.ListedColormap([ '#FFFFFF', '#08C5F5', '#0091F3', '#3898FF', '#008243', '#00A433', '#00D100', '#01F508', '#77FF00', '#E0D100', '#FFDC01', '#EEB200', '#F08100', '#F00101', '#E20200', '#B40466', '#ED02F0' ]) norm = BoundaryNorm(levels, ncolors=cmap.N, clip=True) # Defining the crs crs = area_def.to_cartopy_crs() # Generating a timelist for every hour timelist = [ (basetime + pd.Timedelta(minutes=60*i-10)) for i in range(4) ] # Obtaining the slice of the xarray to be plotted da_plot = reflectivity.sel(time=timelist) # Defining motion quivers qx = motion_u.coords['easting'].values[::5] qy = motion_u.coords['northing'].values[::5] qu = motion_u.values[::5, ::5] qv = motion_v.values[::5, ::5] # Defining coastlines hires = cfeature.GSHHSFeature( levels=[1], scale='h', edgecolor='k' ) # Plotting p = da_plot.plot( col='time', col_wrap=2, subplot_kws={'projection': crs}, cbar_kwargs={ 'extend': 'max', 'ticks': levels[1:], 'format': '%.3g' }, cmap=cmap, norm=norm ) for idx, ax in enumerate(p.axes.flat): ax.quiver(qx, qy, qu, qv, pivot='mid', regrid_shape=20) ax.add_feature(hires) # coastlines ax.gridlines() ax.set_title( "Reflectivity\n" f"Based @ {basetime.strftime('%H:%MH')}", loc='left', fontsize=9 ) ax.set_title( '' ) ax.set_title( f"{basetime.strftime('%Y-%m-%d')} \n" f"Valid @ {timelist[idx].strftime('%H:%MH')} ", loc='right', fontsize=9 ) plt.savefig( THIS_DIR + f"/../tests/outputs/rover-output-map-mn.png", dpi=300 ) radar_image_time = pd.Timestamp.now() .. GENERATED FROM PYTHON SOURCE LINES 242-252 Accumulating hourly rainfall for 3-hour forecast ------------------------------------------------ Hourly accumulated rainfall is calculated every 30 minutes, the first endtime is the basetime i.e. T+0min. #. Convert reflectivity in dBZ to rainrates in mm/h with dbz2rr(). #. Changing time coordinates of xarray from start time to endtime. #. Convert rainrates to rainfalls in 10 mins with rr2rf(). #. Accumulate hourly rainfall every 30 minutes using multiple_acc(). .. GENERATED FROM PYTHON SOURCE LINES 252-274 .. code-block:: default # Convert reflectivity to rainrates rainrates = dbz2rr(reflectivity, a=300, b=1.4) # Converting the coordinates of xarray from start to endtime rainrates_endtime = rainrates.copy() rainrates_endtime.coords['time'] = \ [ pd.Timestamp(t) + pd.Timedelta(minutes=10) for t in rainrates_endtime.coords['time'].values ] # Convert rainrates to accumulated rainfalls every 10 minutes with rr2rf(). rainfalls = rr2rf(rainrates_endtime, scan_duration=10) # Accumulate hourly rainfall every 30 minutes acc_rf = multiple_acc( rainfalls, basetime, basetime+pd.Timedelta(hours=3) ) acc_time = pd.Timestamp.now() .. GENERATED FROM PYTHON SOURCE LINES 275-282 Plotting rainfall maps --------------------------------------- Define the colour scheme and format and plot using xarray.plot(). In this example, only hourly images will be plotted. .. GENERATED FROM PYTHON SOURCE LINES 282-357 .. code-block:: default # Defining the colour scheme levels = [ 0, 0.5, 2, 5, 10, 20, 30, 40, 50, 70, 100, 150, 200, 300, 400, 500, 600, 700 ] cmap = matplotlib.colors.ListedColormap([ '#ffffff', '#9bf7f7', '#00ffff', '#00d5cc', '#00bd3d', '#2fd646', '#9de843', '#ffdd41', '#ffac33', '#ff621e', '#d23211', '#9d0063', '#e300ae', '#ff00ce', '#ff57da', '#ff8de6', '#ffe4fd' ]) norm = BoundaryNorm(levels, ncolors=cmap.N, clip=True) # Defining projection crs = area_def.to_cartopy_crs() # Defining zoom extent r = 64000 proj_site = xarray1.proj_site zoom = ( proj_site[0]-r, proj_site[0]+r, proj_site[1]-r, proj_site[1]+r ) # (x0, x1, y0, y1) # Defining times for plotting timelist = [basetime + pd.Timedelta(hours=i) for i in range(4)] # Obtaining xarray slice to be plotted da_plot = acc_rf.sel( easting=slice(zoom[0], zoom[1]), northing=slice(zoom[3], zoom[2]), time=timelist ) # Plotting p = da_plot.plot( col='time', col_wrap=2, subplot_kws={'projection': crs}, cbar_kwargs={ 'extend': 'max', 'ticks': levels, 'format': '%.3g' }, cmap=cmap, norm=norm ) for idx, ax in enumerate(p.axes.flat): ax.add_feature(hires) # using GSHHS coastlines defined previously ax.gridlines() ax.set_xlim(zoom[0], zoom[1]) ax.set_ylim(zoom[2], zoom[3]) ax.set_title( "Past Hour Rainfall\n" f"Based @ {basetime.strftime('%H:%MH')}", loc='left', fontsize=8 ) ax.set_title( '' ) ax.set_title( f"{basetime.strftime('%Y-%m-%d')} \n" f"Valid @ {timelist[idx].strftime('%H:%MH')} ", loc='right', fontsize=8 ) plt.savefig( THIS_DIR + f"/../tests/outputs/rainfall_mn.png", dpi=300 ) rf_image_time = pd.Timestamp.now() .. GENERATED FROM PYTHON SOURCE LINES 358-370 Extract the rainfall values at a specified location ------------------------------------------------------------------ In this example, the rainfall values at the location is assumed to be the same as the nearest gridpoint. 1. Read information regarding the rain gauge stations into a pandas.DataFrame. 2. Extract the rainfall values at the nearest gridpoint to location for given timesteps (in this example, 30 minute intervals). 3. Store rainfall values over time in a pandas.DataFrame. 4. Plot the time series of rainfall at different stations. .. GENERATED FROM PYTHON SOURCE LINES 370-424 .. code-block:: default # Getting rain gauge station coordinates df = pd.read_csv( os.path.join(THIS_DIR, "../tests/samples/manila_rg_list.csv"), delim_whitespace=True, usecols=[0, 3, 4] ) # Extract rainfall values at gridpoint closest to the # location specified for given timesteps and storing it # in pandas.DataFrame. rf_time = [] for time in acc_rf.coords['time'].values: rf = [] for index, row in df.iterrows(): rf.append(acc_rf.sel( time=time, northing=row[2], easting=row[1], method='nearest' ).values) rf_time.append(rf) rf_time = np.array(rf_time) station_rf = pd.DataFrame( data=rf_time, columns=df.iloc[:, 0], index=pd.Index( acc_rf.coords['time'].values, name='time' ) ) print(station_rf) loc_stn = \ ['BAA', 'BUM', 'PAF', 'QUL', 'ZAP', 'ZAA'] loc_stn_drop = [ stn for stn in station_rf.columns.to_list() if stn not in loc_stn ] df_loc = station_rf.drop(loc_stn_drop, axis=1) print(df_loc) # Plotting time series graph for selected stations ax = df_loc.plot(title="Time Series of Hourly Accumulated Rainfall", grid=True) ax.set_ylabel("Hourly Accumulated Rainfall (mm)") plt.savefig(THIS_DIR+"/../tests/outputs/qpf_time_series.png") extract_time = pd.Timestamp.now() .. GENERATED FROM PYTHON SOURCE LINES 425-428 Checking run time of each component -------------------------------------------------------------------- .. GENERATED FROM PYTHON SOURCE LINES 428-447 .. code-block:: default print(f"Start time: {start_time}") print(f"Initialising time: {initialising_time}") print(f"Rover time: {rover_time}") print(f"SLA time: {sla_time}") print(f"Concatenating time: {concat_time}") print(f"Plotting radar image time: {radar_image_time}") print(f"Accumulating rainfall time: {acc_time}") print(f"Plotting rainfall map time: {rf_image_time}") print(f"Extracting and plotting time series time: {extract_time}") print(f"Time to initialise: {initialising_time-start_time}") print(f"Time to run rover: {rover_time-initialising_time}") print(f"Time to perform SLA: {sla_time-rover_time}") print(f"Time to concatenate xarrays: {concat_time - sla_time}") print(f"Time to plot radar image: {radar_image_time - concat_time}") print(f"Time to accumulate rainfall: {acc_time - radar_image_time}") print(f"Time to plot rainfall maps: {rf_image_time-acc_time}") print(f"Time to extract and plot time series: {extract_time-rf_image_time}") .. rst-class:: sphx-glr-timing **Total running time of the script:** ( 0 minutes 0.003 seconds) .. _sphx_glr_download_auto_examples_qpf_manila.py: .. only:: html .. container:: sphx-glr-footer sphx-glr-footer-example .. container:: sphx-glr-download sphx-glr-download-python :download:`Download Python source code: qpf_manila.py ` .. container:: sphx-glr-download sphx-glr-download-jupyter :download:`Download Jupyter notebook: qpf_manila.ipynb ` .. only:: html .. rst-class:: sphx-glr-signature `Gallery generated by Sphinx-Gallery `_