MRMS/HalfDegGauge
Comparisons of rainrates from gauges and MRMS over HalfDeg Gauge Network, MD-VA
Gauge data
A network of 21 individual gauge sites was deployed
in an approximately 0.5-degree grid area over Delmarva Peninsula near Maryland-Virgina boundary.
Three tipping-bucket gauges were equipped at each of Sites 01-09,18,19,29,30 and two at each of Sites 10-17.
Here is a list of these gauge sites and their locations as well as their deployment dates.
SiteID Lat(N) Lon(W) Installation Date
0001 38.03231 75.39695 Fri, 21 February 2014
0002 38.33289 75.70567 Wed, 26 February 2014
0003 38.19720 75.69918 Thu, 29 December 2011
0004 38.16560 75.39945 Thu, 05 January 2012
0005 37.93716 75.46634 Fri, 20 August 2004 (Gauges A&B),Gauge C was installed on Fri, 30 May 2008
0006 37.92781 75.72054 Fri, 16 December 2011
0007 38.01686 75.82676 Thu, 22 December 2011
0008 38.34092 75.51331 Tue, 24 August 2004 (Gauges A&B),Gauge C was installen on Tue, 03 June 2008
0009 38.19685 75.36974 Mon, 24 March 2014, In addition,
NASA38 (A&B) was moved to Site09 on 06 May 2015. So, 5 gauges at Site09.
0010 38.36982 75.65482 Wed, 04 Jun 2014 (NASA0012)
0011 38.01403 75.52278 Mon, 21 July 2014 (NASA0019)
0012 38.26268 75.34166 Thu, 21 Aug 2014 (NASA0025),moved from (37.93456 -75.47101) on 21 Aug 2014,
then moved to Site0016 (38.05473, -75.40990) on 2015-Apr-14
0013 38.33173 75.40291 Thu, 16 Oct 2014 (NASA0022)
0014 38.12071 75.73399 Thu, 05 Feb 2015 (NASA0005)
0015 38.012383 75.455125 Mon, 09 Mar 2015 (NASA0017)
0016 38.054728 75.409900 Tue, 14 Apr 2015 (NASA0025)
0017 38.36986 75.31886 Mon, 08 Jun 2015 (NASA0006)
0018 38.09558 75.43140 Thu, 11 Feb 2016
0019 38.21044 75.58657 Wed, 09 Mar 2016
0029 37.97574 75.50856 Wed, 25 Aug 2016
0030 37.88065 75.43800 Tue, 05 Jan 2016, data valid after Thu, 23 Feb 2017
All these gauges measure rainfall in increments of 0.01 inch (0.254 mm).
Collocated gauge comparisons
Comparisons with MRMS
The following figure shows the areal map of HalfDegGrid Gauge Networks along with 0.1o
grids. The dots denote locations of the individual gauge sites with their identification
numbers indicated on the right.
The rain tip data from each gauge are QCed, and compared with the
MRMS rainrates at each 0.01-degree pixel where the gauge
locates. The comparisons are conducted at hourly and daily
resolutions for each matched pixel, as well as for averaged pixel
from all matched data.
Two sets of MRMS files, hourly and half-hourly are available for
this study. Hourly MRMS files are available from 01 Feb 2013 to
06 Jun 2016, and half-hourly files from 01 Jun 2014 to present
with a lof of missings (eg. Jan 01 - Jun 07, 2016; Sep 01-23, 2016
and many other periods).
Below is the comparison using hourly MRMS data (Mar 2014 - Dec 2015)
- Scatter plot
- Mean difference and relative error
- PDF by occurrence and volume
In the PDF plot, rainrate thresholds are set to 0.1, 0.2 and 0.5 mm/h.
The inserted "Agreement" in the PDF plot is defined as the fraction of both MRMS and gauge report
the same rain category: rain ( rainrate >threshold), no rain (rainrate <=threshold) among all available
observations.
Agreement=N[(r>T & g>T) + (r<=T & g<=T)] / N(r>=T & g>=T)
where T is rain threshold (0.1,0.2 0r 0.5 mm/h),
N is sample size, r is MRMS rainrate, and g is gauge rainrate.
Be definition, Agreement is between 0 (worst) and 1 (best).
Below is the comparison using QCed half-hourly MRMS files (Jun
2014 - Aug 2019)
The half-hourly MRMS data are QCed using following filters, and then
accumulated to hourly and daily rainrates.
RQI>=80; 0.5<=ratio<=2. and ratio=0.; liquid precipitation only
- Cumulative rainfall
- Scatter plot
- Mean difference and relative error
- PDF by occurrence and volume
|
