Abstract

Using in situ data, three precipitation classes are identified for rainy seasons of West and East Africa: weak convective rainfall (WCR), strong convective rainfall (SCR), and mesoscale convective systems (MCSs). Nearly 75% of the total seasonal precipitation is produced by the SCR and MCSs, even though they represent only 8% of the rain events. Rain events in East Africa tend to have a longer duration and lower intensity than in West Africa, reflecting different characteristics of the SCR and MCS events in these two regions. Surface heating seems to be the primary convection trigger for the SCR, particularly in East Africa, whereas the WCR requires a dynamical trigger such as low-level convergence. The data are used to evaluate the performance of the recently launched Integrated Multisatellite Retrievals for Global Precipitation Measurement (IMERG) project. The IMERG-based precipitation shows significant improvement over its predecessor, the Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis (TMPA), particularly in capturing the MCSs, due to its improved temporal resolution.

1. Introduction

Nearly all tropical precipitation is a result of moist convection that generates both convective and stratiform clouds (Houze 1997; Trapp 2013). Based on various criteria including size, radar reflectivity, and structure of the clouds, the rain-producing systems are grouped into different classes ranging from shallow single-cell convection to mesoscale convective systems (MCSs) (Mohr and Zipser 1996; Tokay and Short 1996; Houze 2004). The MCSs, squall lines, and smaller clusters of deep convection produce most of the rainfall in tropical Africa (Mohr et al. 1999; Friesen 2002; Nesbitt et al. 2006; Jackson et al. 2009). In addition to their direct local environmental impacts, these systems can modulate rainfall processes in other regions within and beyond Africa (Kiladis et al. 2006; Laing et al. 2011; Mekonnen and Thorncroft 2016). Despite their crucial socioenvironmental consequences, the characteristics of rainfall events in tropical Africa such as their intensity and duration are not well understood, partly due to the lack of in situ data. Previous studies are primarily based on the data provided by the Tropical Rainfall Measuring Mission (TRMM) satellite. However, because of limitations in data resolution and uncertainties in infrared (IR)-based cloud retrieval algorithms, the TRMM alone cannot offer a full understanding of the rain events, most of which have a much shorter duration than the satellite-resolved temporal resolution. Potential improvements may be achieved by using the recently launched Global Precipitation Measurement (GPM) satellite that offers a higher resolution and better radar sensitivity than its predecessor, TRMM (Hou et al. 2014; Skofronick-Jackson et al. 2017). Although a number of studies have evaluated the GPM-based products over various regions (e.g., Sharifi et al. 2016; Tan et al. 2016; Tang et al. 2016), such literature is limited for Africa because of the lack of ground-based data (Hill et al. 2016; Sahlu et al. 2016).

The objective of this analysis is twofold. First, the very high-frequency precipitation records from a recently installed network of weather stations in West and East Africa are analyzed. This enables us to provide a gauge-based rainfall classification, find the contribution of each type to the total rainfall, make comparisons between West and East Africa, and identify the surface atmospheric conditions favorable for convection initiation. The second goal is to use these data to evaluate the precipitation product of the Integrated Multisatellite Retrievals for GPM (IMERG) project in the region and to examine its capability to capture the development and propagation of the MCSs.

2. Data and methods

The lack of in situ observations has hindered climate studies in Africa. The recently initiated Trans-African Hydro-Meteorological Observatory (TAHMO) project aims to tackle this issue by installing 20 000 low-cost weather stations throughout the continent in the coming years (van de Giesen et al. 2014). These stations, partly sponsored by the NASA Goddard Space Flight Center, report measurements of the standard meteorological variables in near–real time at 5-min intervals. This study analyzes the precipitation characteristics of four TAHMO stations in West Africa and seven stations in East Africa (Fig. 1a). The study focuses on the late rainy season of each region during 2015, that is, September–November (SON) and November–December (ND) for West and East Africa, respectively (Fig. 1b). The choice of the study period is determined by the availability of data that have undergone a rigorous quality control process. The rain events of each station are identified as the consecutive intervals with rainfall rate greater than 0.1 mm h−1. Using the information from the corresponding stations in each region, two sets of regional rain events are compiled. These events are then classified based on their duration and rainfall rate, using Ward’s hierarchical clustering technique (Ward 1963). The 2-m air temperature T and relative humidity (RH) of different rainfall classes are examined to infer the association between surface conditions and convection initiation.

Fig. 1.

(a) Location of the TAHMO weather stations. (b) Annual cycle of rainfall (mm) during 2015, using TMPA data for West (solid line) and East (dashed line) Africa. The months used are marked with orange squares and circles for West and East Africa, respectively. (c) Three precipitation classes identified for West Africa based on duration and rainfall rate of all rain events during SON. (d) WCR, (e) SCR, and (f) MCS: for each class, the black line is the mean rainfall rate of all events, whose number is presented by the brown line, and the shading shows plus/minus one-half std dev. (g)–(j) As in (c)–(f), but for East Africa during ND.

Fig. 1.

(a) Location of the TAHMO weather stations. (b) Annual cycle of rainfall (mm) during 2015, using TMPA data for West (solid line) and East (dashed line) Africa. The months used are marked with orange squares and circles for West and East Africa, respectively. (c) Three precipitation classes identified for West Africa based on duration and rainfall rate of all rain events during SON. (d) WCR, (e) SCR, and (f) MCS: for each class, the black line is the mean rainfall rate of all events, whose number is presented by the brown line, and the shading shows plus/minus one-half std dev. (g)–(j) As in (c)–(f), but for East Africa during ND.

Frequency analysis is performed to compare the TAHMO data to the TRMM Multisatellite Precipitation Analysis (TMPA) 3B42, version 7 (V7), and the “Final Run” precipitation of IMERG, version 04A (V04A), both of which are gauge calibrated. This is important because IMERG will replace TMPA by mid-2018, when its retrospective processing over the TRMM era is planned to be complete (Huffman et al. 2015). We anticipate that the half-hourly interval of the IMERG data should capture the MCS more accurately than the 3-hourly interval that the TMPA data provide.

3. Precipitation classes

Using the TAHMO data in West Africa, 1122 rain events are identified during the SON season (Fig. 1c). These events are grouped into three different classes based on the magnitude of their intensity and duration. The first class (Fig. 1d), which includes the majority of the events (94%), represents weak convective rainfall (WCR) and contributes 27.1% of the total rainfall in SON. The second class (Fig. 1e) represents strong convective rainfall (SCR) events that reach their peak intensity after about 20 min of their onset and typically last less than 80 min. Although only 3% of the rain events are of this type, the SCR accounts for 47% of the total rainfall in this season. The third class (Fig. 1f) has a longer duration and a moderate peak rainfall rate that resembles the MCSs. These systems, which also include about 3% of all events, produce 25.7% of the seasonal rainfall in West Africa.

Analysis of the precipitation data from the seven stations in East Africa yielded 2202 rain events for ND 2015 (Fig. 1g). These events are clustered based on rainfall rate and duration, providing a three-type classification with temporal evolution similar to that in West Africa (Figs. 1g–j). Most events (91.7%) fall into the WCR class, supplying 23.3% of the seasonal rainfall. The remaining events that occur as SCR and MCS account for 31.5% and 45.3% of the rainfall, respectively. The MCSs in this region have a longer duration and a smaller peak intensity than in West Africa.

Although the mean rainfall rate of the two regions is quite comparable, differences are noticeable in their 75th and 95th percentiles, for which East Africa has, respectively, a higher and lower value than West Africa (Fig. 2a). Their 50th and 90th percentiles have nearly the same values. The larger magnitude of the extreme rainfall rates in West Africa can be attributed to the SCR events, as can be seen by comparing Figs. 1c and 1g. The rain events in East Africa, however, have a slightly longer mean duration than in West Africa (Fig. 2b). This is also apparent in the 75th–95th percentiles, and particularly in the durations longer than 200 min that represent the MCSs.

Fig. 2.

Violin plots presenting the PDF of (a) rainfall rate and (b) duration of all rain events in West and East Africa during SON and ND, respectively. These are based on the TAHMO (gauge), IMERG, and TMPA datasets. The number of events identified by each data/region is shown above the upper x axis. The 50th, 75th, 90th, and 95th percentiles are also shown.

Fig. 2.

Violin plots presenting the PDF of (a) rainfall rate and (b) duration of all rain events in West and East Africa during SON and ND, respectively. These are based on the TAHMO (gauge), IMERG, and TMPA datasets. The number of events identified by each data/region is shown above the upper x axis. The 50th, 75th, 90th, and 95th percentiles are also shown.

4. Surface conditions and convective development

To understand the surface atmospheric conditions associated with the three different precipitation classes, T and RH from the TAHMO stations are examined. We acknowledge that surface meteorological conditions alone cannot provide a full picture of convective initiation and growth. However, a three-dimensional suite of meteorological data at the time scale relevant to convection is not available for the region. This study, therefore, uses the TAHMO’s high-temporal-resolution observations to infer useful information with the available data. The mean surface conditions for 3 h prior to the onset of the WCR and SCR events are examined (Fig. 3). For the MCSs that have a long duration and are most likely advected from other regions, T and RH are evaluated during the lifetime of the events. The mean conditions for the no-rain intervals and the entire period are also provided in Table 1. Both regions have a high mean RH for their respective season. The mean temperature, however, is much higher in West Africa, mainly because of its lower altitude.

Fig. 3.

Box plots of temperature (solid boxes) and relative humidity (dotted boxes) associated with the rain events of the three precipitation classes identified in Fig. 1 for (a) West and (b) East Africa. The left and right y axes represent T and RH, respectively. For WCR and SCR, mean surface conditions of these two variables are used between 3 h prior and the onset of the events. For MCS, these are evaluated during the lifetime of the events. Note that the colors used here correspond to those in Figs. 1c–j.

Fig. 3.

Box plots of temperature (solid boxes) and relative humidity (dotted boxes) associated with the rain events of the three precipitation classes identified in Fig. 1 for (a) West and (b) East Africa. The left and right y axes represent T and RH, respectively. For WCR and SCR, mean surface conditions of these two variables are used between 3 h prior and the onset of the events. For MCS, these are evaluated during the lifetime of the events. Note that the colors used here correspond to those in Figs. 1c–j.

Table 1.

The mean surface temperature and relative humidity of the no-rain intervals and the entire rainy season for West and East Africa. Because of the large range of their duration and the fact that some dry spells can last several days, these intervals are divided into three categories: 1–3, 3–6, and greater than 6 h. The intervals shorter than 1 h are counted as noise and ignored. Depending on the length of each category, the surface conditions are evaluated simultaneously or at 3-h lag to perform a consistent comparison with the rain events.

The mean surface temperature and relative humidity of the no-rain intervals and the entire rainy season for West and East Africa. Because of the large range of their duration and the fact that some dry spells can last several days, these intervals are divided into three categories: 1–3, 3–6, and greater than 6 h. The intervals shorter than 1 h are counted as noise and ignored. Depending on the length of each category, the surface conditions are evaluated simultaneously or at 3-h lag to perform a consistent comparison with the rain events.
The mean surface temperature and relative humidity of the no-rain intervals and the entire rainy season for West and East Africa. Because of the large range of their duration and the fact that some dry spells can last several days, these intervals are divided into three categories: 1–3, 3–6, and greater than 6 h. The intervals shorter than 1 h are counted as noise and ignored. Depending on the length of each category, the surface conditions are evaluated simultaneously or at 3-h lag to perform a consistent comparison with the rain events.

The surface temperature prior to the onset of the WCR events is about 3°C below the average for both regions (Fig. 3). The mean RH for these events is about 94%, which represents nearly saturated air at the surface. Given some weak to moderate low-level convergence, air at this RH can quickly reach its condensation level, resulting in a low cloud-base height. These clouds would be shallow because of the weak buoyancy, given the parcels’ relatively cool initial virtual temperatures.

Comparing the no-rain intervals with rain events enables us to identify the favorable conditions for rain-producing systems. Based on their duration, the no-rain periods are divided into three categories. Depending on the length of each category, the surface conditions are evaluated either simultaneously or at 3-h lags to perform a consistent comparison with the results shown in Fig. 3. The 3-h lag mean T and RH of no-rain intervals with 1–3 h duration (Table 1) are very similar to those observed for the WCR. We hypothesize that it does not rain during these periods because of the absence of dynamical triggers vigorous enough for the parcels to reach their lifting condensation level (LCL). We cannot confirm this because of the lack of high-resolution wind data.

The surface conditions during the SCR events are quite different from those for the WCR. Over the 3-h period prior to these events, the data in both regions show an enhanced temperature at the surface, implying that the surface heating plays an important role in convection initiation of the SCR. The majority of these events have an RH ranging between 70% and 90%, with an average of about 80% in both regions. This leads to a relatively higher LCL and cloud-base height than in the WCR events. The higher surface temperatures result in greater parcel buoyancy and boundary layer turbulence such as convective rolls required for deep convection. These conditions are consistent with the high rainfall intensity of the SCR events. This paradigm is particularly important for East Africa, where the positive temperature anomalies are statistically significant at the 8% level, using a Student’s t test. This suggests that boundary layer turbulence resulting from enhanced variable surface heating is the primary trigger of convection in this region. However, the temperature difference is less significant in West Africa, implying that the surface heating is a necessary but not sufficient condition for convection initiation. Formation of deep convection in this region most likely requires other features such as low-level convergence (Dezfuli 2017).

During the long-duration events, the surface experiences cool temperatures relative to the seasonal mean and near-saturation-level RH, and both of these variables show a small range of variation. This occurs because the precipitation enhances the evaporation locally, leading to a temperature drop and an increase in moisture content of the surface air. This information does not provide insight into the triggers of the MCSs; rather, it confirms the surface conditions expected to occur during an MCS.

5. Satellite versus ground-based precipitation

TMPA has been instrumental to success in numerous climate studies, particularly over the data-limited regions like Africa (e.g., Anyah et al. 2006; Naumann et al. 2012; Dezfuli et al. 2015; Ichoku et al. 2016). The IMERG project continues the legacy of the TMPA in monitoring space-based precipitation observations and provides significant advantages in spatiotemporal resolution. Using the 3-hourly TMPA and half-hourly IMERG data, rain events for West and East Africa are identified during their respective rainy seasons. That allows us to perform intercomparisons of frequency analysis between the satellite and gauge-based events (Fig. 2a). Only the rainfall rate is investigated, since most of the gauge-based events have a lifetime shorter than the temporal resolution of the satellites.

As shown in Fig. 2a, IMERG offers some advantages over TMPA in capturing the distribution of intensity of the rainfall events for both regions. In West Africa, the seasonal mean of the rainfall rate obtained from the IMERG events is closer to the gauge records than that of the TMPA. This is also apparent in the extreme events, represented by the values greater than the 90th percentile. These events are primarily of the SCR type. In East Africa, the IMERG substantially outperforms the TMPA, as evident in the mean and the three highest percentiles. The TMPA overall overestimates the rainfall rate in this region. The difference between low-intensity events from the in situ and satellite observations in both regions results from the majority of these events, represented, for example, by the 50th percentile, having a duration shorter than the satellite-resolving time intervals. However, IMERG shows some improvements over TMPA.

To evaluate the added value of higher temporal resolution of IMERG in detecting the MCSs, the time evolution of a sample rain event is examined (Fig. 4). This was the longest rain event in West Africa during the study period. It depicts a westward-propagating MCS where the time evolution and spatial structure of its rainfall intensity is consistent with the information drawn from the gauge records. The propagation speed of this MCS is about 13 m s−1. A possible explanation for this may lie in the interaction between the cold pool and the background winds, schematically shown in Fig. 4q.

Fig. 4.

(a)–(p) Time evolution of the longest MCS in West Africa during SON 2015, captured by IMERG. Each panel shows a half-hourly snapshot of the IMERG Final Run precipitation of IMERG V04A, starting at 2230 UTC 20 Oct 2015. Date/time of each interval is shown in lower-right part of its corresponding panel. The event occurs between 0200 and 0515 UTC 21 Oct 2015, and its start time is presented in (h). The small red box shows the area where the TAHMO weather stations are located. (q) A schematic of a possible mechanism responsible for the westward propagation of this MCS is also presented. This relies on interaction between the westerly vertical shear generated by the cold pool and the easterly shear driven by the dominant regional winds: the low-level westerly (LLW) and African easterly jet (AEJ), which during SON are present over the equatorial latitudes (Fink and Reiner 2003; Nicholson and Grist 2003; Dezfuli 2017).

Fig. 4.

(a)–(p) Time evolution of the longest MCS in West Africa during SON 2015, captured by IMERG. Each panel shows a half-hourly snapshot of the IMERG Final Run precipitation of IMERG V04A, starting at 2230 UTC 20 Oct 2015. Date/time of each interval is shown in lower-right part of its corresponding panel. The event occurs between 0200 and 0515 UTC 21 Oct 2015, and its start time is presented in (h). The small red box shows the area where the TAHMO weather stations are located. (q) A schematic of a possible mechanism responsible for the westward propagation of this MCS is also presented. This relies on interaction between the westerly vertical shear generated by the cold pool and the easterly shear driven by the dominant regional winds: the low-level westerly (LLW) and African easterly jet (AEJ), which during SON are present over the equatorial latitudes (Fink and Reiner 2003; Nicholson and Grist 2003; Dezfuli 2017).

6. Discussion and conclusions

In situ surface weather data recorded at 5-min intervals have been analyzed to better understand various precipitation characteristics in tropical West and East Africa. The rain events are divided into three classes, broadly defined as

  • WCR: duration <40 min and intensity <10 mm h−1,

  • SCR: duration <80 min and intensity >10 mm h−1, and

  • MCS: duration >80 min and intensity <10 mm h−1.

The SCR and MCS have a combined contribution of approximately 75% to the total seasonal rainfall, although they represent only about 8% of the rain events. Despite general similarities in both regions, rain events in East Africa tend to have a longer duration and lower intensity than in West Africa. Surface heating appears to be the primary convection trigger for the SCR in East Africa, when RH is about 80%, whereas the SCR in West Africa seems to require an additional mechanical lifting mechanism such as low-level convergence. The WCR, however, would emerge when a dynamical trigger is present and the surface air is nearly saturated.

The gauge observations have allowed us to validate the TMPA and IMERG precipitation estimates for this data-limited region. IMERG shows improvements over TMPA in capturing precipitation characteristics. In addition, IMERG captures the propagation of large MCSs in equatorial Africa due to its half-hourly time intervals, a significant advantage over the TMPA’s 3-hourly temporal resolution, which misses the time evolution of most of these systems. Although this finding was based on the Final Run gauge-calibrated data (Fig. 4), similar patterns were found using other IMERG products, such as the IR-only, Early Run, and Late Run (not shown). This suggests strong potential for near-real-time analysis of these systems based on IMERG.

The findings presented here strictly apply only to the geographic and temporal scope of this study. Some uncertainties may arise from the fact that our analysis focuses on select seasons/regions, compares point gauge data with gridded products, and does not fully investigate the atmospheric dynamics. This study serves as a basis for additional research on these issues. As the IMERG product continues providing new data and the TAHMO station network grows, we will be able to perform interseasonal and interannual analysis in many more parts of Africa in the near future. The new data would also allow us to investigate a large pool of MCS cases, offering a comprehensive understanding of their characteristics.

Acknowledgments

We thank the two anonymous reviewers for their constructive remarks. A.K.D.’s research was supported by the NASA Postdoctoral Program (NPP) at the Goddard Space Flight Center, administrated by the Universities Space Research Association (USRA) through a contract with NASA. This research was also supported under the NASA Research Opportunities in Space and Earth Sciences (ROSES)–2009 and 2013 Interdisciplinary Studies (IDS) Program (Dr. Jack Kaye, Earth Science Research Director), Grant NNH12ZDA001N-IDS, through the Radiation Sciences Program managed by Dr. Hal Maring.

REFERENCES

REFERENCES
Anyah
,
R. O.
,
F. H.
Semazzi
, and
L.
Xie
,
2006
:
Simulated physical mechanisms associated with climate variability over Lake Victoria basin in East Africa
.
Mon. Wea. Rev.
,
134
,
3588
3609
, doi:.
Dezfuli
,
A. K.
,
2017
: Climate of western and central equatorial Africa. Oxford Research Encyclopedia of Climate Science, 66 pp., doi:.
Dezfuli
,
A. K.
,
B. F.
Zaitchik
, and
A.
Gnanadesikan
,
2015
:
Regional atmospheric circulation and rainfall variability in south equatorial Africa
.
J. Climate
,
28
,
809
818
, doi:.
Fink
,
A. H.
, and
A.
Reiner
,
2003
:
Spatiotemporal variability of the relation between African easterly waves and West African squall lines in 1998 and 1999
.
J. Geophys. Res.
,
108
, 4332, doi:.
Friesen
,
J.
,
2002
: Spatio-temporal rainfall patterns in Northern Ghana. Diploma thesis, Geographic Institute, Rheinische Friedrich-Wilhelm-Universität Bonn, 81 pp.
Hill
,
P. G.
,
R. P.
Allan
,
J. C.
Chiu
, and
T. H.
Stein
,
2016
:
A multisatellite climatology of clouds, radiation, and precipitation in southern West Africa and comparison to climate models
.
J. Geophys. Res. Atmos.
,
121
,
10 857
10 879
, doi:.
Hou
,
A. Y.
, and Coauthors
,
2014
:
The Global Precipitation Measurement mission
.
Bull. Amer. Meteor. Soc.
,
95
,
701
722
, doi:.
Houze
,
R. A.
, Jr.
,
1997
:
Stratiform precipitation in regions of convection: A meteorological paradox?
Bull. Amer. Meteor. Soc.
,
78
,
2179
2196
, doi:.
Houze
,
R. A.
, Jr.
,
2004
:
Mesoscale convective systems
.
Rev. Geophys.
,
42
,
RG4003
, doi:.
Huffman
,
G. J.
,
D. T.
Bolvin
, and
E. J.
Nelkin
,
2015
: Integrated Multi-satellitE Retrievals for GPM (IMERG) technical documentation. NASA/GSFC Code 612 Tech. Doc., 48 pp. [Available online at http://pmm.nasa.gov/sites/default/files/document_files/IMERG_doc.pdf.]
Ichoku
,
C.
, and Coauthors
,
2016
:
Biomass burning, land-cover change, and the hydrological cycle in northern sub-Saharan Africa
.
Environ. Res. Lett.
,
11
,
095005
, doi:.
Jackson
,
B.
,
S. E.
Nicholson
, and
D.
Klotter
,
2009
:
Mesoscale convective systems over western equatorial Africa and their relationship to large-scale circulation
.
Mon. Wea. Rev.
,
137
,
1272
1294
, doi:.
Kiladis
,
G. N.
,
C. D.
Thorncroft
, and
N. M.
Hall
,
2006
:
Three-dimensional structure and dynamics of African easterly waves. Part I: Observations
.
J. Atmos. Sci.
,
63
,
2212
2230
, doi:.
Laing
,
A. G.
,
R. E.
Carbone
, and
V.
Levizzani
,
2011
:
Cycles and propagation of deep convection over equatorial Africa
.
Mon. Wea. Rev.
,
139
,
2832
2853
, doi:.
Mekonnen
,
A.
, and
C. D.
Thorncroft
,
2016
:
On mechanisms that determine synoptic time scale convection over East Africa
.
Int. J. Climatol.
,
36
,
4045
4057
, doi:.
Mohr
,
K. I.
, and
E. J.
Zipser
,
1996
:
Defining mesoscale convective systems by their 85-GHz ice-scattering signatures
.
Bull. Amer. Meteor. Soc.
,
77
,
1179
1189
, doi:.
Mohr
,
K. I.
,
J. S.
Famiglietti
, and
E. J.
Zipser
,
1999
:
The contribution to tropical rainfall with respect to convective system type, size, and intensity estimated from the 85-GHz ice-scattering signature
.
J. Appl. Meteor.
,
38
,
596
606
, doi:.
Naumann
,
G.
,
P.
Barbosa
,
H.
Carrao
,
A.
Singleton
, and
J.
Vogt
,
2012
:
Monitoring drought conditions and their uncertainties in Africa using TRMM data
.
J. Appl. Meteor. Climatol.
,
51
,
1867
1874
, doi:.
Nesbitt
,
S. W.
,
R.
Cifelli
, and
S. A.
Rutledge
,
2006
:
Storm morphology and rainfall characteristics of TRMM precipitation features
.
Mon. Wea. Rev.
,
134
,
2702
2721
, doi:.
Nicholson
,
S. E.
, and
J. P.
Grist
,
2003
:
The seasonal evolution of the atmospheric circulation over West Africa and equatorial Africa
.
J. Climate
,
16
,
1013
1030
, doi:.
Sahlu
,
D.
,
E. I.
Nikolopoulos
,
S. A.
Moges
,
E. N.
Anagnostou
, and
D.
Hailu
,
2016
:
First Evaluation of the Day-1 IMERG over the Upper Blue Nile basin
.
J. Hydrometeor.
,
17
,
2875
2882
, doi:.
Sharifi
,
E.
,
R.
Steinacker
, and
B.
Saghafian
,
2016
:
Assessment of GPM-IMERG and other precipitation products against gauge data under different topographic and climatic conditions in Iran: Preliminary results
.
Remote Sens.
,
8
(
2
),
135
, doi:.
Skofronick-Jackson
, and Coauthors
,
2017
:
The Global Precipitation Measurement (GPM) mission for science and society
.
Bull. Amer. Meteor. Soc.
, doi:, in press.
Tan
,
J.
,
W. A.
Petersen
, and
A.
Tokay
,
2016
:
A novel approach to identify sources of errors in IMERG for GPM ground validation
.
J. Hydrometeor.
,
17
,
2477
2491
, doi:.
Tang
,
G.
,
Y.
Ma
,
D.
Long
,
L.
Zhong
, and
Y.
Hong
,
2016
:
Evaluation of GPM Day-1 IMERG and TMPA version-7 legacy products over mainland China at multiple spatiotemporal scales
.
J. Hydrol.
,
533
,
152
167
, doi:.
Tokay
,
A.
, and
D. A.
Short
,
1996
:
Evidence from tropical raindrop spectra of the origin of rain from stratiform versus convective clouds
.
J. Appl. Meteor.
,
35
,
355
371
, doi:.
Trapp
,
R. J.
,
2013
: Mesoscale-Convective Processes in the Atmosphere. Cambridge University Press, 377 pp.
van de Giesen
,
N.
,
R.
Hut
, and
J.
Selker
,
2014
:
The Trans-African Hydro-Meteorological Observatory (TAHMO)
.
Wiley Interdiscip. Rev.: Water
,
1
,
341
348
, doi:.
Ward
,
J. H.
, Jr.
,
1963
:
Hierarchical grouping to optimize an objective function
.
J. Amer. Stat. Assoc.
,
58
,
236
244
, doi:.

Footnotes

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).