THE QUALITY OF H-SAF ATOVS BASED
PRECIPITATION PRODUCTS OVER POLAND
Bozena Łapeta, and Danuta Serafin-Rek
Satellite Remote Sensing Department, Institute of Meteorology and Water Management
National Research Institute | 14 P. Borowego Str. | 30-215 Kraków, Poland
INTRODUCTION
H-SAF CONSORTIUM
The main objectives of H-SAF are:
The main goal of EUMETSAT Satellite Application Facility
in Support to Operational Hydrology and Water Management (H-SAF) is to provide satellite products for operational hydrology. Products of H-SAF concerns precipitation,
soil moisture and snow cover parameters. Among them,
the H-SAF precipitation products based on both passive
microwave sensors (AMSU and MSH) and IR sensors calibrated by MW have been operationally available for meteorological and hydrological users.
a. t o provide new satellite-derived products from existing and
future satellites with sufficient time and space resolution to satisfy
the needs of operational hydrology; identified products:
• precipitation (liquid, solid, rate, accumulated);
• s oil moisture (at large-scale, at local-scale, at surface, in the
roots region);
• s now parameters (detection, cover, melting conditions, water
equivalent);
In the poster, the quality of mentioned above H-SAF
precipitation products in precipitation detection and estimation is analysed. The analysis was performed using
data from both rain gauges and radar data from Polish
ground measurement networks. The quality of the satellite products was studied using continuous and categorical statistical parameters within appropriate precipitation classes.
b. to perform independent validation of the usefulness of the new
products for fighting against floods, landslides, avalanches, and
evaluating water resources; the activity includes:
ownscaling/upscaling modelling from observed/predicted
•d
fields to basin level;
• f usion of satellite-derived measurements with data from radar
and raingauge networks;
• a ssimilation of satellite-derived products in hydrological models;
• a ssessment of the impact of the new satellite-derived products
on hydrological applications.
COMPOSITION OF THE H-SAF CONSORTIUM
No. Country
Units in the Country (responsible unit in bold)
Role in the Project
01
Austria
- Zentral Anstalt für Meteorologie und Geodynamik
- Technische Univ. Wien, Inst. Photogrammetrie & Fernerkundung
Leader for soil moisture
02
Belgium
- Institut Royal Météorologique
03
Bulgary
- National Institute of Meteorology and Hydrology
04
ECMWF
- European Centre for Medium-range Weather Forecasts
Contributor for “core” soil moisture
05
Finland
- Finnish Meteorological Institute
Leader for snow parameters
06
France
- Météo-France
- CNRS Centre d’Etudes Spatiales de la BIOsphere
- CNRS Centre d’études des Environnem. Terrestres et Planétaires
07
Germany
- Bundesanstalt für Gewässerkunde
08
Hungary
- Hungarian Meteorological Service
09
Italy
-
10
Poland
- Institute of Meteorology and Water Management
11
Romania
- National Meteorological Administration
11
Slovakia
- Slovenský Hydrometeorologický Ústav
13
Turkey
-
More information at H-SAF web page: www.hsaf.meteoam.it
Servizio Meteorologico dell’Aeronautica
Dipartimento Protezione Civile, Presidenza Consiglio Ministri
CNR Istituto di Scienze dell’Atmosfera e del Clima
Ferrara University, Department of Physics
Turkish State Meteorological Service
Middle East Technical University, Civil Engineering Department
Istanbul Technical University, Meteorological Department
Anadolu University
Host + Leader for precipitation
Leader for Hydrology
Contributor for “core” snow
parameters
H-SAF ATOVS BASED PRECIPITATION PRODUCT
H-02 - Precipitation rate at ground by MW
cross-track scanners (with indication of phase)
All 15-km pixels with brightness temperatures at 183±7 GHz that are below a threshold T7 are flagged as potentially
precipitating, where
Product H-02 is based on the instruments AMSU-A and
AMSU-B or MHS flown on NOAA and MetOp satellites.
The algorithm has been devoloped in CNR-ISAC.
T7 = 0.667 (T53.6 - 248) + 252 + 6 cos θ
and where θ is the satellite zenith angle and T53.6 is the spatially filtered limb-corrected 53.6 GHz brightness temperature
obtained by selecting the warmest brightness temperature within a 7×7 array of AMSU-B pixels. This T7 threshold was
determined empirically and can vary with atmospheric temperature.
Algorithm:
The 183±7 GHz channel can become sensitive to surface variations in very cold, dry atmospheric conditions.
When T53.6 is less than 248 K, the 183±3 GHz brightness temperature is compared to a threshold T3.
The algorithm uses data from MW absorption bands exploited for temperature sounding (the 54 GHz band of AMSU-A) or for water vapour sounding (the 183 GHz band of
AMSU-B and MHS), in which the effect of surface emissivity is minimized. Precipitation is retrieved by exploiting the
differential effect of liquid drops or ice particles at different
frequencies associated to weighting functions peaking in
different atmospheric layers. It is a highly indirect principle that implies that only part of the retrieval process is
physically-based, whereas substantial part of the retrieval
is currently relying on the use of neural networks trained
with the use of Cloud Resolving Model (CRM).
T3 = 242.5 + 5 cos θ
The thresholds T7 and T3 are slightly colder than a saturated atmosphere would be, implying the presence of a
microwave absorbing or scattering cloud. It is possible for the 183±3 GHz and the 183±1 GHz channels to be sensitive to
surface variations. If T53.6 is less than 242 K, then the pixel is assumed not to be precipitating.
Summary H-02 performance:
• Status: Operational.
• Coverage: Strips of ~ 2250 km swath crossing the H-SAF area [25-75°N lat, 25°W-45°E long] in direction approx.
S-N or N-S
• Cycle: Up to six passes/day (if one Met-Op-A and two NOAA satellites are available) at approximately 09:30, 01:30,
03:00 ECT (descending mode) and 21:30, 13:30, 15:00 ECT (ascending mode)
Flow chart of the AMSU-MHS (H-02) precipitation rate processing chain.
• Spatial Resolution: Corresponds to the nominal resolution of MHS, varying with the viewing scan angle
from 16 x 16 km2 / circular at nadir to 26 x 52 km2 / ovate at scan edge
• Accuracy:
PRECIPITATION RANGE
THRESHOLD
TARGET
OPTIMAL
> 10 mm/h
90
80
25
1-10 mm/h
120
105
50
< 1 mm/h
240
145
90
Accuracy requirements for product PR-OBS-2 [RMSE (%)]
PRECIPITATION RATE
• Timeliness: 30 min from observing time
• Dissemination: By dedicated lines to centres connected by GTS - By EUMETCast to most other users,
especially scientific
• Formats: BUFR with values on grid points corresponding to the MHS orbital projection.
Also JPEG or similar for quick-look
H-02 VALIDATION RESULTS: JUNE 2013-JUNE 2014
PDF Spring
Mean Error, rr>=0.25 mm/h
RD
1,0
PDF Summer
H-02
RD
0,6
URD, precipitation classes
400%
H-02
[0.25,1)
10000000
100000
0,8
CONCLUSIONS
[1,10)
>=10
350%
1000000
10000
300%
0,4
• The H-02 precipitation products underestimates the
measured rain rate for the whole period, however,
the slight overestimation was found for light precipitation during summer 2013.
100000
0,2
250%
1000
10000
0,0
200%
1000
100
-0,2
150%
100
10
PDF Autumn
Mean Error, precipitation classes
2
RD
29≤ PR <30
28≤ PR <29
27≤ PR <28
26≤ PR <27
25≤ PR <26
24≤ PR <25
23≤ PR <24
22≤ PR <23
21≤ PR <22
20≤ PR <21
19≤ PR <20
18≤ PR <19
17≤ PR <18
16≤ PR <17
15≤ PR <16
14≤ PR <15
13≤ PR <14
12≤ PR <13
11≤ PR <12
10≤ PR <11
8≤ PR <9
9≤ PR <10
7≤ PR <8
6≤ PR <7
5≤ PR <6
3≤ PR<4
4≤ PR <5
0.25≤ PR <1
29≤ PR <30
28≤ PR <29
27≤ PR <28
26≤ PR <27
25≤ PR <26
24≤ PR <25
23≤ PR <24
22≤ PR <23
21≤ PR <22
20≤ PR <21
19≤ PR <20
18≤ PR <19
17≤ PR <18
16≤ PR <17
15≤ PR <16
14≤ PR <15
Jun
2014
13≤ PR <14
May
2014
12≤ PR <13
Apr.
2014
11≤ PR <12
Mar.
2014
10≤ PR <11
Feb.
2014
8≤ PR <9
Jan.
2014
9≤ PR <10
Dec.
2013
7≤ PR <8
Nov.
2013
6≤ PR <7
Oct.
2013
5≤ PR <6
Sep.
2013
3≤ PR<4
Aug.
2013
4≤ PR <5
Jul.
2013
2≤ PR<3
Jun.
2013
1≤ PR<2
-1,0
50%
1
0.25≤ PR <1
1
2≤ PR<3
-0,8
100%
10
-0,6
1≤ PR<2
-0,4
0%
Jun. Jul. 2013 Aug.
2013
2013
Sep.
2013
Oct.
2013
RD
100000
Nov.
2013
POD
PDF Winter
H-02
Dec.
2013
Jan.
2014
Feb.
2014
Mar.
2014
Apr.
2014
May Jun 2014
2014
10000
0,9
0
0,8
10000
-2
-4
1000
0,7
0,6
1000
0,5
100
-6
100
0,4
-8
0,3
10
10
0,2
-10
REFERENCES
Algorithms Theoretical Baseline Document for product H02 – PR-OBS-2, 2011, at www.hsaf.meteoam.it
29≤ PR <30
28≤ PR <29
27≤ PR <28
26≤ PR <27
25≤ PR <26
24≤ PR <25
23≤ PR <24
22≤ PR <23
21≤ PR <22
20≤ PR <21
19≤ PR <20
18≤ PR <19
17≤ PR <18
16≤ PR <17
15≤ PR <16
14≤ PR <15
13≤ PR <14
12≤ PR <13
11≤ PR <12
10≤ PR <11
9≤ PR <10
8≤ PR <9
7≤ PR <8
6≤ PR <7
5≤ PR <6
4≤ PR <5
3≤ PR<4
2≤ PR<3
1≤ PR<2
0.25≤ PR <1
29≤ PR <30
28≤ PR <29
27≤ PR <28
26≤ PR <27
25≤ PR <26
24≤ PR <25
23≤ PR <24
22≤ PR <23
21≤ PR <22
20≤ PR <21
19≤ PR <20
18≤ PR <19
17≤ PR <18
16≤ PR <17
15≤ PR <16
14≤ PR <15
13≤ PR <14
12≤ PR <13
11≤ PR <12
May Jun 2014
2014
10≤ PR <11
Apr.
2014
9≤ PR <10
Mar.
2014
8≤ PR <9
>=10
Feb.
2014
7≤ PR <8
[1,10)
Jan.
2014
6≤ PR <7
[0.25,1)
Dec.
2013
5≤ PR <6
Nov.
2013
4≤ PR <5
Oct.
2013
3≤ PR<4
Sep.
2013
2≤ PR<3
Jun. Jul. 2013 Aug.
2013
2013
0,1
1
1≤ PR<2
1
0.25≤ PR <1
-12
0
Jun.
2013
Jul.
2013
Aug.
2013
Sep.
2013
Oct.
2013
Nov.
2013
Dec.
2013
• The quality of H-02 rain rate products depends on
the season – the best results were obtained for autumn and winter and worst – for spring and summer. This dependence is strongly pronounced for
light precipitation class.
• Analysis performed for different precipitation classes showed that the quality of considered satellite
precipitation product increases with increasing rain
rate. The worst results were obtained for light precipitation – the URD reached 350% when compared
with radar data, respectively.
FAR
1
H-02
The H-SAF TOVS based precipitation product (H-02)
was validated with the use of radar data for the period of June 2013-June 2014. On the basis of presented
material it can be stated that:
Jan.
2014
Feb.
2014
Mar.
2014
Apr.
2014
May
2014
Jun
2014
• The ability of H-SAF rain rate products in precipitation
recognition was found to be rather poor for winter
and autumn and reasonably good for summer.