Preparing Meteorological Data

Jump to: navigation, search


In the following two methods to prepare meteorological data are discussed.

  1. Using Observations that you already have
  2. Using the WATCH dataset
    1. Downscaling precipitation data

Using UMEP

Location in UMEP

This plugin can be used to transform temporal meteorological data into the specific format used in UMEP. The format used is a space separated format where the time related variables are separated into year, day of year, hour and minute. However, the plugin is able to process other time formats such as month, day of month etc.

Plugin location in UMEP

From the UMEP plugin select the MetPreprocessor

  • UMEP
    • Pre-processor
      • Meteolorological data

Running the plugin

Interface for inputting an asci data file into the correct format for SUEWS

When you run the plugin, you will see the dialog shown on the right. It consists of four sections.

  1. The top left section lets you select an existing text file including meteorological data of any time resolution. However, at least hourly resolution is recommended. The data can separated by a number of general separators and include any number of header lines.
  2. The middle left section lets you specify time related columns of your imported data file.
  3. The right section lets you choose the columns of your imported data file that corresponds with the meteorological variables used in UMEP.
  4. The lower left section lets is where you perform the export, close the plugin etc.

The variables currently included are:

No. Header name Description
1 iy Year [YYYY]
2 id Day of year [DOY]
3 it Hour [H]
4 imin Minute [M]
5 qn Net all-wave radiation [W m-2]
6 qh Sensible heat flux [W m-2]
7 qe Latent heat flux [W m-2]
8 qs Storage heat flux [W m-2]
9 qf Anthropogenic heat flux [W m-2]
10 U Wind speed [m s-1]
11 RH Relative Humidity [%]
12 Tair Air temperature [°C]
13 pres Surface barometric pressure [kPa]
14 rain Rainfall [mm]
15 kdown Incoming shortwave radiation [W m-2]
16 snow Snow [mm]
17 ldown Incoming longwave radiation [W m-2]
18 fcld Cloud fraction [tenths]
19 Wuh External water use [m3]
20 xsmd (observed) soil moisture [m3 m-3 or kg kg-1]
21 lai (observed) Leaf area index [m2 m-2]
22 kdiff Diffuse radiation [W m-2]
23 kdir Direct radiation [W m-2]
24 wdir Wind direction [°]

Perform quality control (recommended)

Tick this in if you want to perform a simple quality control where the plugin search for unreasonable values for the specific variables that you want to include. If a variable is included, no values out of range can be included.

The button Export data will first provide you with a window that let you specify the location of your output file. Then the export process starts.

The Close button closes the plugin.

Using WATCH data


The WATCH data can be downloaded for free with any of the following approaches:

  • URL (e.g. for easy checking of directory structure, update dates and file sizes): and click on /WATCH_Forcing_Data and /WFDEI

  • ftp downloads of individual files:, un=rfdata, pw=forceDATA then: “cwd /WFDEI”.

  • Python script:

use the Python script provided by us.

By running, you will be able to specify the download path on your local machine, variable(s) and date range to download.

The downloaded files will be put in the same hierarchy as stored in the WATCH server.

Note: if interpolation with our Python wrapper will be conducted afterwards, please DO keep the hierarchy as specified by this script.


Downscaling precipitation data

    • This section moved here from SUEWS manual - needs checking!**


  • This is not included in the current version of the model - please ignore this section. (useRainModel must not appear in RunControl currently)
  • RainParameters.nml specifies empirical parameters for rainfall a disaggregation model when useRainModel=1 in RunControl.nml.
  • If useRainModel=0 in RunControl.nml, then this file does not need to be present.
  • When useRainModel=1, a time-varying rainfall sequence is generated at the model time resolution using a stochastic process.
  • When useRainModel=0, the rainfall accumulations are spread evenly across the model time steps.

In both cases the accumulated rainfall occurring within each bin of the input rainfall data is conserved.

The input parameters for the downscaler are derived from fits to empirical data and should therefore be updated to represent the location studied. The values provided are for London, UK, based on three years of Radar data (2012-2014 inclusive).

Name Definition Units Value Site
rainIntensityLogMean Geometric mean of a log-normal distribution describing the rain intensity.
  • specified in log-space, i.e. log(geometric mean)
mm h-1 -1.103937465 London, UK (2012)
rainIntensityLogSd Log-standard deviation of a log-normal distribution describing the rain intensity
  • specified in log-space, i.e. log(geometric sd)
mm h-1 0.365991286 London, UK (2012)
rainExtentScale Scale parameter of a Weibull distribution fitted to the CDF of temporal extent (start of first rain pulse to end of final rain pulse) of rainfall clusters.
  • cluster:a series of rain pulses separated by no more than X minutes, where X is 180 min in the example dataset.
min 218.93547799 London, UK (2012)
rainExtentShape Shape parameter of a Weibull distribution fitted to the CDF of temporal extent of rainfall clusters as defined above. 218.93547799 London, UK (2012)
rainExponent Exponent of a power law function fitted to the empirical power spectrum of the rainfall time series in the frequency interval (0.001, 0.1] min-1
rainWetProb Proportion of cluster time occupied by actual rainfall: i.e. total time spent raining ÷ total duration of rain clusters 0.5236637 London, UK (2012)


People who have contributed to this

Name Comments
Lingbo XUE University of Reading Python code
Ting SUN University of Reading Python code, supervision, WATCH, HW
Helen Ward University of Reading Radiation, Testing, Supervision, downscaling precipitation
Andy GABEY University of Reading Radiation, Testing
Zhe ZHANG University of Reading Initial codes
Tom Kokkonen University of Helsinki Preciptitation
Fredrik Lindberg University of Gothenburg Main wrapper for SUEWS
Yin San TAN University of Reading Precipitation
Leena Jarvi University Helsinki
Sue Grimmond University of Reading


  • Kokkonen et al. (2016, in review)
  • Ward et al. (2016, in review)
  • Gabey et al. (2016, in review)


  • Tan YS (2015) MSc, University of Reading
  • Xue, L (2016) In preparation, MSc, University of Reading
  • Zhang Z (2015) Undergraduate, University of Reading