Technical ReportPDF Available

Evaluation and comparison of satellite-based rainfall product CHIRPS and reanalysis product ERA5 in West Africa

Authors:
Atmosphere-Climate-Continental Landmass Program
Master in Earth, Planetary and Environmental Sciences
Université Grenoble Alpes
Evaluation and comparison of satellite-based rainfall
product CHIRPS and reanalysis product ERA5 in
West Africa
Dinh Ngoc Thuy Vy
Research project performed at Institute of Environmental Geosciences
Under the supervision of:
Thierry Lebel
Chagnaud Guillaume
Defended before a jury composed of:
Defense date, xx/xx/2020
ii
Declaration
I hereby declare that the present master’s thesis was composed by myself and that the work
contained herein is my own. I also confirm that I have only used the specified resources. All
formulations and concepts taken verbatim or in substance from printed or unprinted material
or from the Internet have been cited according to the rules of good scientific practice and
indicated by footnotes or other exact references to the original source.
The present thesis has not been submitted to another university for the award of an academic
degree in this form. This thesis has been submitted in printed and electronic form. I hereby
confirm that the content of the digital version is the same as in the printed version.
I understand that the provision of incorrect information may have legal consequences.
Dinh Ngoc Thuy Vy
iii
Acknowledgement
First of all, I have the pleasure to express my gratitude to Prof. Thierry Lebel for giving me the
opportunity to have this internship. Many thanks for his guidance and supervision for helping
me complete this report.
My sincere gratitude for Guillaume Quantin and Chagnaud Guillaume for their help in collecting
the necessary data and various information to make this report.
Finally, I would like to thank Flavie Constantin and Carméline Meli who helped me in handling
administrative procedures, so that I could start my internship.
iv
Abstract
Rainwater is a valuable resource for the agricultural countries in West Africa, but the global
climate change is threatening rainfed agriculture in these countries. The available rain gauge
network is limited which prevents the efficient water management as well as climatic study in
West Africa. This study aims to evaluate and compare the accuracy of two satellite-based
rainfall products (CHIRPS) and a reanalysis (ERA5) with rain gauge data of the BADOPLU
data base, assembled by the Phyrev team of IGE and documenting in situ rainfall in West
Africa from 1981 to 2015.
The two products provide accurately temporal and spatial precipitation in comparison with
BADOPLU. They have good correlation, especially for the seasonal (in rainy season) and
annual resolution (r > 0.8). The performance of the two products is best over the Sahelian
region, a semi-arid region with a marked dry season alternating with a 3-5 months rainy
season. Over the Soudanian region the transitional region between Sahel and Guinea, both
CHIRPS and ERA5 also perform well at annual scale. The correlation is lower in the Guinean
region which is characterized by a more complex terrain (r = 0.4 to 0.6). CHIRPS gives more
accurate rainfall estimation than ERA5 in general. ERA5 outperforms CHIRPS only in the east
coast of West Africa in dry season.
v
Résumé
L'eau de pluie est une ressource précieuse pour les pays agricoles d'Afrique occidentale, mais
le changement climatique mondial menace l'agriculture pluviale dans ces pays. Le réseau de
pluviomètres disponible est limité, et le recours aux produits satellitaux est donc souvent
indispensable pour faire un bon suivi des campagnes agricoles et servir d’entrée aux modèles
hydrologiques. Cette étude vise à évaluer sur cette région un produit pluviométrique
satellitaires (CHIRPS), ainsi que le produit pluviométrique des réanalyses ERA5, qui sert de
référence pour de nombreuses études diagnostiques ou de modélisation rétrospective. Cette
évaluation se fait en utilisant les cumuls pluviométriques journaliers de la base de données
BADOPLU comme référence, assemblée par l’équipe Phyrev de l’IGE avec l’aide des services
météorologiques nationaux et qui couvre l’Afrique de l'Ouest de 1982 à 2015.
Le produit satellitaire CHIRPS et le produit de réanalyse ERA5 présentent une bonne
corrélation avec le produit interpolé à partir des données sol de BADOPLU aux échelles
saisonnière et annuelle (r > 0.8). La performance des deux produits est meilleure dans la
région du Sahel, c'est la région qui divise le plus clairement les saisons (presque pas de pluie
en saison sèche). En zone Soudanienne - la région de transition entre le Sahel et la Guinée,
les deux produits fournissent également de bonnes estimations des précipitations annuelles
et mensuelles. La corrélation est plus faible en Guinée qui possède un terrain complexe, côtier
au Nord et montagneux à des extrémités Est et Ouest (r = 0.4 to 0.6). CHIRPS donne une
estimation des précipitations plus précise que ERA5 en général. ERA5 ne surpasse CHIRPS
que sur la côte est de l'Afrique de l'Ouest aux échelles spatiales et temporelles considérées.
vi
Table of contents
Declaration ..................................................................................................................................ii
Acknowledgement ...................................................................................................................... iii
Abstract ..................................................................................................................................... iv
Résumé ...................................................................................................................................... v
Table of contents ....................................................................................................................... vi
1 Introduction ......................................................................................................................... 1
1.1 General context ........................................................................................................... 1
1.2 The rainfall data issue in West Africa .......................................................................... 1
1.3 Objectives of the study ................................................................................................ 2
2 Data and method ................................................................................................................ 3
2.1 Data.............................................................................................................................. 3
2.1.1 Satellite product ........................................................................................................ 3
2.1.2 Reanalysis product .................................................................................................... 3
2.1.3 Rain gauge ................................................................................................................ 3
2.1.4. Data pre-processing ................................................................................................. 4
2.2 Study domains and period ........................................................................................... 5
2.3 Statistical analysis ....................................................................................................... 6
3 Results ................................................................................................................................ 7
3.1 Statistics analysis between ERA5 and CHIRPS ......................................................... 7
3.2 Evaluation precipitation of ERA5, CHIRPS and BADOPLU ..................................... 10
3.2.1. Evaluation of annual precipitation estimation ........................................................ 10
3.2.2. Evaluation of seasonal precipitation estimation .................................................... 11
3.2.3. Evaluation of monthly precipitation estimation ...................................................... 12
4 Discussion ........................................................................................................................ 13
Conclusion and perspective ..................................................................................................... 14
Reference ................................................................................................................................. 15
Appendix ................................................................................................................................... 17
1
1 Introduction
1.1 General context
Rainwater is a valuable resource for the agricultural countries in West Africa, but global climate
change is threatening rainfed agriculture in these countries (B. Sultan et al. 2013). The region is
known for its vulnerability to rainfall (and more generally climate) variability that impacts the natural
resources (water, vegetation) and subsequently the welfare of the population, in societies where the
economy is based mainly on agriculture (Benjamin Sultan et al. 2010). Hence, reliable rainfall data
in West Africa are of utmost importance to monitor long term changes in the regime of precipitations
and for operational applications linked to agriculture.
The area has many differences in climatic and weather characteristics due to the presence of the
ocean to the West and to the South and of the Sahara Desert to the North. Large scale rainfall
patterns are driven by a monsoon regime, which produces a South to North negative rainfall gradient
(Lebel, Diedhiou, and Laurent 2003).
Rainfall in West Africa has several characteristics e.g: (i) an overall regional pattern of the annual
cycle, characterized by the alternance of a single rainy season with a wet season, except close to
the coast of the Guinea Gulf, where a little dry season is embedded in the rainy season; ii) a South
to North negative rainfall gradient, associated both to a reduction of the length of the rainy season
and to a stronger time intermittency when moving to the North; iii) rainfall being essentially of
convective origin, it is highly variable in space and time at the event scale and heavy rainfall in a
short time often causes serious soil erosion and flooding, despite an overall shortage of water.
The central objective of this work is thus to undertake a preliminary investigation of whether two
rainfall products as different in nature as CHIRPS and ERA5 are, can retrieve these main
characteristics and how accurately they do it.
1.2 The rainfall data issue in West Africa
The ground-based rain-gauge networks are generally scarce and have been degrading over the last
few decades (Nicholson et al. 2003a). In addition, the complex topography of West Africa also
hindered the rational distribution and expansion of rain gauge networks in the region (Akinsanola et
al. 2017). According to the World Meteorological Organization (WMO), a good rainfall monitoring
over sub-Saharan Africa would require a uniform distribution of at least 3000 stations (ideally
10,000); however, only 744 stations are in operation (Satgé et al. 2020). The available gauge
network is thus limited by many spatial and temporal gaps which prevent efficient water management
in West Africa (Satgé et al. 2020).
Satellite rainfall estimates are increasingly and widely used because they provide a rainfall
monitoring in nearly real time with a complete spatial coverage, which is not the case with ground-
based networks. Validated data from satellites thus contribute to fill gaps associated with rain gauge
networks. Following the global trend, the validation of various satellite-based rainfall products has
early been studied in West Africa for TRMM (Nicholson et al. 2003b), GPCP (Ali et al. 2005), CPC
(Pierre et al. 2011). More recently a few works undertook a comparison between several products.
Gosset et al. (2013) evaluated seven satellite-based rainfall products (PERSIANN, CMORPH,
TMPA-RT v.6, TMPA-Adj v.6, GSMaP-MVK, GCPC-1dd, and RFE-2) at daily/1o resolution, over the
rainy season (JuneSeptember) between the years 2003 and 2010 for two densely instrumented
sites of the AMMA-CATCH observing system in Niger and Benin. Dembélé et al. (2016) evaluated
seven satellite-based rainfall data sets (ARC 2.0, CHIRPS, PERSIANN, RFE 2.0, TARCAT, TRMM
3B42, and TRMM 3B43) with rain gauge data between 2001 and 2014. CHIRPS was found to be
correlated well in the territory of Burkina. Satgé et al. (2020) compared 23 gridded precipitation
datasets across West Africa with rain gauges measurement at the daily and monthly time scales for
a limited period of 4 years (20002003).
It turns out from Dembélé et al. (2016) and from Satgé et al. 2020 that CHIRPS can provide reliable
rainfall estimates of monthly precipitation in the region, with more contrasted results at the daily
scale. At any rate, a comprehensive validation over a long period and over the whole region remains
2
to be carried out. Besides, in West Africa, some of studies were often confined to the country or
basin scale and/or over a short period of a few years. Indeed, the performance of the satellite rainfall
products was not always the same for all regions and not for the entire period. Depending on the
region (especially topography), this may affect the accuracy of different satellite rainfall products
(Derin and Yilmaz 2014; Gao and Liu 2013). In addition, the recent long-term evaluation and
comparison of satellite rainfall data and rain gauge data is essential because of variable performance
of satellites over years (Tang et al. 2020). Because a comprehensive evaluation of rainfall satellite
products is still missing for the region, the goal of this study was to work on a sufficiently long period
to make sure that different climatological conditions are covered.
On the other hand, the ERA-5 reanalysis (Hersbach et al. 2020) provides rainfall data at the hourly
timescale. However, ERA-5 does not incorporate any direct rainfall measurement be it from ground
sensors or satellites. Rainfall is produced as the result of the internal dynamics of a global circulation
model, forced by reanalyzed observations of wind, temperature and humidity. By nature it is thus
less accurate than a satellite product and expected to provide appropriate rainfall estimates at the
global scale (Tang et al. 2020).
Since the time available for this study was limited, it was chosen to focus on two different types of
products: CHIRPS is a satellite product, with the longest period covered (since 1981), which allows
for (more) robust climatological studies; and the other is the ERA-5 reanalysis product, which is not
used in an operational mode but is the most widely used for diagnostic studies.
1.3 Objectives of the study
The evaluation of the CHIRPS and ERA-5 products are carried out by using the rain gauge data of
the BADOPLU data base, assembled by the Phyrev team of IGE and documenting in situ rainfall in
West Africa from 1981 to 2015. Two specific questions are to be addressed:
1. The difference between the two rainfall products ERA5 and CHIRPS in terms of climatology
(mean precipitation over each of the three sub-regions considered) and meteorology (annual,
seasonal and monthly precipitation); ideally the daily scale was also targeted, but it proved
unfeasible to address that scale in such a small period of work.
2. The accuracy of these two products in comparison to BADOPLU.
While CHIRPS is rated to be highly accurate in precipitation estimates, ERA5 gives a high
appreciation of precipitation in terms of temporal resolution (i.e., hourly precipitation), which makes
a promising tool for diagnostic studies when it comes to linking atmospheric dynamics with rainfall
intensities.
3
2 Data and method
2.1 Data
The study used two products to compare with rain gauge data (BADOPLU): ERA5 (Fifth generation
of ECMWF atmospheric reanalyzes of the global climate) and CHIRPS (Climate Hazards Group
InfraRed Precipitation with Station data). These two products are fundamentally different in their
conception, but they are both characterized by high spatial and temporal resolutions, a long-time
coverage and are easy for download.
Table 1. Description of the characteristics of satellite product and reanalysis product.
Product
Unit
Resolution
(Spatial / Temporal)
Period
Spatial coverage
ERA5
mm/hour
0.25°/1 hour
1979 to present
Global
CHIRPS
m/day
0.05°/1 day
1981 to present
50°S-50°N, 0° 360°E
2.1.1 Satellite product
CHIRPS was developed by the US Geological Survey (USGS) and the Climate Hazards Group at
the University of California, Santa Barbara (UCSB). The main data sources used in the creation of
CHIRPS were: (1) the monthly precipitation climatology; (2) quasi-global geostationary thermal
infrared (IR) satellite observations from two NOAA sources, the Climate Prediction Center (CPC) IR
and the National Climatic Data Center (NCDC) B1 IR; (3) the Tropical Rainfall Measuring Mission
(TRMM) 3B42 product from NASA; (4) atmospheric model rainfall fields from the NOAA Climate
Forecast System, version 2 (CFSv2); and (5) in situ precipitation observations obtained from a
variety of sources including national and regional meteorological services (Funk et al. 2015).
CHIRPS is free to download through the USGS website
(https://data.chc.ucsb.edu/products/CHIRPS-2.0/).
2.1.2 Reanalysis product
ERA5 (Hersbach et al. 2020) provides a description of the recent climate through a combination of
outputs from a Numerical Weather Prediction (NWP) model and observations of key atmospheric
variables (e.g. temperature, humidity, wind). It is the successor of the ERA-interim and the last of a
suite of reanalyzes products (e.g. ERA-15, ERA-40) developed by the European Centre for Medium-
Range Weather Forecasts (ECMWF). The precipitation data are available at the hourly timescale
with a 0.25° (~30km) horizontal resolution and are freely accessible to users from the Climate Data
Store (CDS) website (https://cds.climate.copernicus.eu).
2.1.3 Rain gauge
The BADOPLU (BAse de DOnnées PLUviomètres), database comes from the fusion and update of
various data sources which existed previously (ORSTOM, AGRHYMET). It consists of 53 268
station-years collected from 1981 to 2015 (Figure 1, upper panel). In this study, we used the daily,
spatially interpolated precipitation on a 1°×1° regular grid (e.g. Figure 1, lower panel, the
accumulative annual precipitation) (Panthou et al. 2018) which was used as the reference
precipitation for comparison with the satellite-based and reanalysis-generated rainfall products for
the period 1981 to 2015.
4
Figure 1. Map of location of rain gauge stations (upper panel) and mean annual precipitation (lower
panel) on ground observations (BADOPLU data base) from 1981 to 2015. The grey color in the
upper panel represents the number of operation years of rain gauge station (provided by Guillaume
Quantin 2020).
2.1.4. Data pre-processing
CHIRPS and ERA5 do not have the same spatial and temporal resolution. Therefore, a
preprocessing step to get to the same resolution is required. A Python package, xESMF was applied
to regrid the two evaluated products (https://xesmf.readthedocs.io). It’s simple and easy to use
because it is based on Python's well-established standards (numpy and xarray). In case of
decreasing resolution of satellite images, a regridding bilinear algorithm (linear regridding in two
dimensions) was selected.
BADOPLU is the rain data from gauge stations in the form of points. It thus needs to be interpolated
to get to the same resolution as that of the two evaluated products (ERA5, CHIRPS) for comparison.
The interpolated data in this study was provided by (Chagnaud et al. 2020); it was an optimal kriging
interpolation.
5
Figure 2. Processing stages for evaluation and comparison precipitation of ERA5, CHIRPS and
BADOPLU.
2.2 Study domains and period
The study area was a part of West Africa, located between 0° 20°N and 20°W 20°E (Figure 3).To
accurately and comprehensively analyze the relationship between CHIRPS, ERA5 and the rain
gauge data BADOPLU in the study area from 1981 to 2015, an analysis by climatic regions was
recommended. The climatic zones used in this study were defined as follows: Sahel (11-160N),
Soudanian region (8-110N) and Guinea (4-80N). The alternance of rainy and dry seasons is under
the control of the West African Monsoon dynamics (Nicholson 2013). There is a single rainy season
occurring from May/June to September/October in the Sahel and from April to October in the
Soudanian region with average annual rainfall ranging from 150- 600 mm and 600 - 1200 mm
respectively. In Guinea, the rainy season lasts from late March to October, but is interrupted by a
little dry season in July; the average annual rainfall ranges from 2200 to 5000 mm (USGS 2016).
Figure 3. Map of the study area.
6
2.3 Statistical analysis
After being processed to the same resolution, the comparison was conducted by statistics analysis.
The first was a comparison between CHIRPS and ERA5 to get the correlation between the two
products, the second was a comparison of these two products with BADOPLU. Three temporal
resolutions (mean annual precipitation, seasonal precipitation and monthly precipitation) were
considered, along with three regions: Sahel, Soudanian, Guinea region. The comparison results
were expressed through statistical indicators, including the Pearson correlation coefficient (r), the
root mean square error (RMSE) and the percent bias (pbias) (or relative difference).
(1) The Pearson correlation coefficient (r) (Eq.1) is used to assess the correlation between the two
evaluated products themselves or to compare with the rain gauges data. The Pearson correlation
coefficient defined in (1) ranges from 1 to 1 and gets perfect value at 1.




(Eq.1)
where:
n: number of data pairs;
: gauge rainfall measurement;
: average gauge rainfall measurement;
: satellite rainfall estimate;
: average satellite rainfall estimate.
(2) The root means square error (RMSE) (Eq.2) measures how much difference there is between
two data sets (e.g. rain gauge and satellite product precipitation data). The smaller an RMSE value,
the closer estimated and observed values are. RMSE ranges from 0 to ∞ and gets perfect score at
0.
RMSE nyixi
n
i
(Eq.2)
(3) The percent bias (Eq.3) is the proportion differences between the data of the rain gauges and
ERA5 or CHIRPS. Percent bias ranges from 0 to ∞ and gets perfect score at 0.



(Eq.3)
7
3 Results
Preprocessed data of the two products ERA5, CHIRPS and in situ observations (BADOPLU) with
spatial resolution 10x10 are respectively evaluated in terms of annual, seasonal and monthly
precipitation.
3.1 Statistics analysis between ERA5 and CHIRPS
As can be seen on Figure 4a,b both products display the same overall pattern of rainfall over West
Africa at the annual scale:
Latitude: Rainfall decreases with latitude (0 - 200N), divided into 3 climatic regions with lowest
rainfall in the Sahel (rainfall <500mm / year) and highest in Guinea (rainfall> 1500mm / year).
Longitude: Rainfall is high in areas near the sea and high terrain, the value of annual rainfall
in these regions is always over 1700mm/year, most evident in Guinea. Over the rest of West
Africa (North of ~ 10°N), the climatological mean annual rainfall is much more homogenous.
Figure 4. Comparison of average annual precipitation between ERA5 and CHIRPS over the 35
years of the study period (1981 2015).
The average annual rainfall in period 1981-2015 of ERA5 and CHIRPS shows a similar distribution
of rainfall. However, the mean correlation between two products is only 0.41. Their low correlation
may be derived from the annual rainfall fluctuations of each product. This also proves that one of the
two products may be inaccurate, requiring verification with BADOPLU.
Figure 4c shows the correlation between ERA5 and CHIRPS across the study area. Good correlation
(r > 0.7) is mainly in the Sahel to the North, the other regions having a correlation of about 0.4 to 0.6.
The two products have very low correlation (<0.4) at 150W - 100W (coastal area) and in areas with
high elevation topography (> 800m, see Figure 2).
It is thus striking that in many areas the two products share less than 25% of joint variability (r <0.5).
The correlation of the two products is best over the Sahel region, for reasons that might be both of
dynamical and statistical origin, a matter that needs more investigation. The correlation is lower at
Guinea - the region with almost year-round rainfall and complex terrain. Note also that, when looking
at relative differences (Figure 4d), there is a spatial pattern with CHIRPS precipitation being
substantially larger than ERA5 precipitation in the north and differences being much smaller in the
South: mean relative differences of 0 to 30%, except for mountainous areas in the Soudanian region
(mean relative differences from 75 to 100%).
8
Because general circulation models are known to ill-position the ITCZ during the rainy season (Cook
and Vizy 2006), we can assume that there is a negative bias of precipitation over the Sahel in the
ERA5 product.
Since there is a marked difference in rainfall estimates for each climatic zone in West Africa, the
comparison of two products will be made for each of them (Sahel, Soudanian, Guinea region). Table
2 shows the number of dry and wet days in the three climatic zones of West Africa from 1981 to
2015 (12781 days in total over 35 years). The number of dry days was counted based on days
without rainfall for both products. The remaining days (rainfall > 0 for at least one of the two products)
will be the number of wet days. Results shows that the number of rainy days increases significantly
from the Sahel to Guinea. While the number of rainy days in the Sahel accounts for only about 55%
of the total number of days in 35 years, in Guinea 91% of the days are rainy (Table 2).
Table 2. Number of dry days and wet days in three climatic regions of West Africa from 1981 to 2015.
Number of dry days
(rainfall = 0)
Number of wet days
(rainfall >0)
Sahel
5692 (45%)
7089 (55%)
Soudanian
2993 (23%)
9788 (77%)
Guinea
1099 (9%)
11682 (91%)
The Pearson coefficient of correlation (r) and RMSE are calculated based on wets days of ERA5
and CHIRPS. This eliminates interference correlation from dry days of both products (a series of
zero rainfall).
Figure 5. Time series precipitation in three distinct regions of West Africa from 1981 to 2015
(Different vertical axis scale).
9
Figure 5 shows the time series of daily precipitation in three climatic zones of West Africa. The
intensity of precipitation in West Africa decreases gradually from Guinea to the Sahel. Both CHIRPS
and ERA5 show that the Guinea region is the region with the highest daily rainfall. Daily rainfall in
the Sahel reaches a maximum of about 15 mm/day while in the Soudanian region and Guinea it can
be up to 30 and 40 mm/day, respectively. However, Guinea found a lot of variation during the 35
years of the survey, especially the rainfall estimated from the CHIRPS product. Extreme rainfall
increases in Guinea (in 2008, 2009, 2011, 2014) may originate from the storms that CHIRPS can
capture.
Estimated daily rainfall over the 35 years of the CHIRPS product always tends to be higher than that
of ERA5. The results of correlation analysis (r) show that they only got the highest correlation in
Soudanian (r=0.64) and Sahel (r = 0.6). In Guinea, the correlation of daily rainfall between these two
rainfall products is lower (r = 0.5). Maybe CHIRPS can capture more extreme rain (Figure 5c) while
the highest values of rain by ERA5 are quite similar over 35 years.
Figure 6. Correlation of seasonal precipitation of ERA5 and CHIRPS.
Daily precipitation was aggregated to seasonal accumulation for comparison. The rainy seasons in
Sahel, Soudanian and Guinea are from June to September, from April and October and from March
to October, respectively. The dry season included accumulative precipitation from October to May
in Sahel, from November to March in Soudanian and November to February. Seasonal correlation
is calculated based on the seasonal precipitation of each region in 35 years (i.e. n = 35 in Eq.1)
Correlation of seasonal precipitation in three climatic zones between ERA5 and CHIRPS is around
0.7 in rainy season and from 0.4 to 0.6 in dry season (Figure 6). The lowest correlation can be
observed in Guinea. The above results show that ERA5 and CHIRPS have a higher correlation for
seasonal precipitation than for daily rainfall, a result that could be expected since the scaled
variability at seasonal scale is far smaller than at daily scale.
In summary, there are some differences between the estimated precipitation by ERA5 and CHIRPS
in several regions of West Africa, especially at the coastal, mountain and northern areas. This
suggests that the performance of either of these products (or both) was inaccurate in some areas
10
and in some periods over the 35 years of the survey. The performance of these two products will be
compared with the rain gauge data (BADOPLU) which will be discussed in the next section.
3.2 Evaluation precipitation of ERA5, CHIRPS and BADOPLU
3.2.1. Evaluation of annual precipitation estimation
Figure 7. Comparison of annual precipitation between ERA5, CHIRPS and BADOPLU in period
1981 2015. (a) and (c) are the correlation (n = 35); (b) and (d) are the mean absolute difference
(n = 35).
Figure 7a shows a comparison of annual precipitation between CHIRPS and BADOPLU. CHIRPS
gets highest correlation (r>0.8) with BADOPLU in the central of West Africa; lower correlation (0 to
0.5) in the eastern regions (range from 10 to 200E), the western regions (coastal in Soudanian and
Guinea range 150W to 50W) and in the northern of Sahel. The areas with low correlation between
CHIRPS and BADOPLU are areas with few rain gauges (see Figure 1 for location of rain gauge
stations), except for the eastern mountainous region from 10 200E. Therefore, this are some
inaccuracies of the interpolated observation precipitation due to the shortage of measuring stations.
In general, additional rainfall station data is required to increase the reliability of the study.
Similarly, Figure 7c shows that ERA5 is not highly correlated with BADOPLU in regions with little or
no measuring stations (r = 0 to 0.2). While CHIRPS is not highly correlated in the East region, ERA5
is highly correlated with BADOPLU in the East range from 15 to 200E (r> 0.7). The center of West
Africa (from 50E - 50W) and the coastal area of the Sahel, where there are many rain-gauging
stations, ERA5 and BADOPLU are still not highly correlated (r = 0.2 to 0.4).
The pixels in white are where the correlation value has not been computed (Figure 7a, Figure 7c).
As mentioned in the introduction, the data of measuring stations is gradually disappearing over time
(see number of stations per year in the appendix). There were thus some missing data (null data) in
some stations over 35 years. During the calculation of correlation, the locations (or pixel) with no
rainfall value will be removed to avoid errors.
Figure 7b and d show the absolute difference between CHIRPS and ERA with respect to BADOPLU.
Both products show significant difference in coastal and mountain regions (> 400 mm/year). In the
Sahel and Soudanian, ERA5 has lower rainfall values than BADOPLU (less than 100 to 300
mm/year), whereas CHIRPS has less difference (from 0 to 100 mm/year). Guinea area has complex
terrain with mountains (from 10 to 150E and 15°W to 10°W). Topography being a complicating factor
when it comes to estimate rainfall, it is expected that the two evaluated products display the largest
discrepancies with respect to BADOPLU in this region. In comparison with BADOLU at Guinea,
11
ERA5 has higher difference (~400 mm/year) than CHIRPS (0 to 200mm/year). In coastal and upland
areas, both CHIRPS and ERA5 do not correlate well with the measuring station, however CHIRPS
is less different than ERA5. Therefore, CHIRPS seem more suitable for complex topographic than
ERA5 at the annual scale.
3.2.2. Evaluation of seasonal precipitation estimation
Correlations of seasonal precipitation of the three climatic zones by dry and rainy season between
BADOPLU with two satellite products (ERA5, CHIRPS) are presented in Table 3 (see the scatter
charts in the appendix for more details).
ERA5 correlates best with BADOPLU in the Sahel in the dry season (with r = 0.71, RMSE = 17
mm/season, Bias = 17%) and dry season (r = 0.68, RMSE = 93 mm/season, Bias = 13%). The
seasonal correlation between CHIRPS and BADOPLU in Sahel is almost perfect, with r > 0.9 in both
dry and rainy season. CHIRPS performs not well at Soudanian, r = 0.6 in the rainy season and r =
0.56 in the dry season, but it is still slightly better than ERA5. Guinea does not observe a high
correlation between CHIRPS and BADOPLU in both rainy season (r = 0.51) and dry season (r =
0.33) as at Sahel or Soudanian. Besides, ERA5 is better than CHIRPS in performance at Guinea in
dry season with r = 0.61.
Table 3. Evalution the accuary of dry and rainy season of ERA and CHIRPS.
Season
Indicators
Sahel
Soudanian
Guinea
ERA5
CHIRPS
ERA5
CHIRPS
ERA5
CHIRPS
Rainy season
r
0.68
0.9
0.56
0.6
0.47
0.51
RMSE
(mm/season)
93
37
210
262
312
355
Bias (%)
13
4
13
17
34
38
Dry season
r
0.71
0.92
0.54
0.56
0.61
0.33
RMSE
(mm/season)
17
10
23
17
136
87
Bias (%)
17
10
32
23
83
52
12
3.2.3. Evaluation of monthly precipitation estimation
Figure 8. Comparison of monthly precipitation between ERA5, CHIRPS and BADOPLU in 3
climatic zones from 1981 to 2015.
Figure 8 shows monthly precipitation of ERA5, CHIRPS, and BADOPLU in three climatic zones. In
Sahel, the rainy and dry seasons are clearly divided. Since this is a hot, dry area with low air humidity,
satellite product do well in detecting rain without many confounding factors. In term of monthly
precipitation, ERA5 is systematically smaller than BADOPLU while CHIRPS is approximately equal
to BADOPLU.
Although the precipitation is negligible in dry season at Sahel and to a lesser extent Soudanian, both
ERA5 and CHIRPS can still satisfyingly detect the precipitation at these regions. Unlike the Sahel,
the Soudanian area has more vegetation. The humidity is higher than in Sahel, which can affect the
rain detection of CHIRPS and rainfall calculation of ERA5 in Soudanian, especially during the rainy
months. This suggests that in this region there are more factors that interfere with the accuracy of
CHIRPS and ERA5 than in Sahel.
In Guinea, CHIRPS and ERA5 display larger monthly precipitation amounts than BADOPLU. ERA5
precipitation values tend to be higher than CHIRPS, BADOPLU mostly in the dry months, while
CHIRPS tends to be higher in the rainy months. This may prove that ERA5 can be disturbed during
the dry season.
In summary, the results of statistical comparison between the two satellites ERA5 and CHIRPS show
that they have low correlation (r in range 0.4 to 0.6) in the description of rain areas in West Africa,
especially in mountains or coastal areas. The estimated rainfall intensity all shows a gradual increase
in rainfall southward of West Africa. In addition, both products show rainfall with two distinct seasons
in West Africa. The results of evaluating the accuracy of the two products compared with the
observed value (BADOPLU) show that CHIRPS is best correlated with the observation in Sahel
(r>0.9 in 2 seasons). CHIRPS is more correlated than ERA5 in many areas, while ERA5 gives better
results in eastern West Africa.
13
4 Discussion
As discussed in the introduction, satellite and reanalysis products for precipitation estimation can
provide a valuable tool in rainwater prediction and management for flood forecasting, water
resources management and agricultural activities in West Africa. However, these products need to
be assessed for their accuracy through measurement data. The data used here for that purpose are
those from the BADOPLU database, which is one of the most complete data sets in this region. This
study used this data set to evaluate the accuracy of ERA5 and CHIRPS between 1981 and 2015 at
a spatial resolution of 1 degree and over a range of timescales. Although the results show a high
correlation, up to 0.9 according to annual or seasonal resolution in some regions, the results are not
as good as coastal or high mountains. These results can be attributed to two reasons: (i) satellite
product inefficiency in complex terrain, (ii) BADOPLU product with inaccurate interpolation in areas
with few measuring stations. In addition, the measuring stations also have measurement quality that
changes over time, leading to inaccuracies in measurement. Besides that, the analyzed areas should
be divided according to topographical characteristics, climate, not just spatial coordinates in order to
improve the accuracy of spatial analysis.
Although errors may exist in comparison between ERA5, CHIRPS and interpolated observation
precipitation (BADOPLU), this study gives similar statements on the performance of CHIRPS in the
West Africa with other studies. A comparison between 23 estimated precipitation products from 2000
- 2003 in West Africa, CHIRPS was one of the best products provide the reliable daily and monthly
precipitation estimates (Satgé et al. 2020). In another study in the period from 2001 to 2014, CHIRPS
was also the satellite with the least error of the seven satellite rainfall estimates in West Africa. In
addition, CHIRPS showed the highest correlation efficiency in rainy seasons (Dembélé and Zwart
2016). This statement is completely true for the performance of CHIRPS in the wet season from
1981 to 2015.
14
Conclusion and perspective
ERA5 and CHIRPS satellite products show quite accurately the distribution by temporal and spatial
precipitation in comparison with BADOPLU observation data (Rainfall varies between climatic
zones). Both products describe the same trend of the spatial and temporal rainfall characteristics of
West Africa. The performance of two products is best in the Sahel region. In the Soudanian region
the transitional region between Sahel and Guinea, both products also perform well from monthly
to annual precipitation estimation. The correlation is lower in the Guinea region which is
characterized by a complex terrain, coastal in the northward and mountains in eastward.
The results of the comparison with the interpolated rain gauge data show that CHIRPS correlates
better with BADOPLU than does ERA5. It can be concluded that, CHIRPS is a product closer to
reality than ERA5. However, in some areas where CHIRPS does not perform well, ERA5 correlates
well with BADOPLU (in the eastern part of coastal West Africa). Therefore, it is preferable to consider
the selection of products for each region in West Africa.
In areas with many rain gauges, CHIRPS correlates well with seasonal/annual precipitation of
BADOPLU (r> 0.8). In areas with few rain gauges stations, the correlation between CHIRPS and
BADOPLU is very low (r <0.5). The correlation result would probably change if more station data
were available. There should be more stations in the eastern regions (range from 5 to 200E), the
western regions (coastal in Soudanian and Guinea - range 150W to 50W) and in the northern of
Sahel. But overall, CHIRPS is a promising satellite product that can be used for research in West
Africa.
15
Reference
Akinsanola, A. A., K. O. Ogunjobi, V. O. Ajayi, E. A. Adefisan, J. A. Omotosho, and S. Sanogo. 2017.
Comparison of Five Gridded Precipitation Products at Climatological Scales over West Africa.
Meteorology and Atmospheric Physics 129 (6): 66989. https://doi.org/10.1007/s00703-016-
0493-6.
Ali, Abdou, Thierry Lebel, and Abou Amani. 2005. Rainfall Estimation in the Sahel. Part I: Error
Function. Journal of Applied Meteorology 44 (11): 16911706.
https://doi.org/10.1175/JAM2304.1.
Chagnaud, Guillaume, Hubert Gallée, Thierry Lebel, Gérémy Panthou, and Théo Vischel. 2020. A
Boundary Forcing Sensitivity Analysis of the West African Monsoon Simulated by the Modele
Atmospherique Regional. Atmosphere 11 (2). https://doi.org/10.3390/atmos11020191.
Cook, Kerry H., and Edward K. Vizy. 2006. Coupled Model Simulations of the West African Monsoon
System: Twentieth- and Twenty-First-Century Simulations. Journal of Climate 19 (15): 3681
3703. https://doi.org/10.1175/JCLI3814.1.
Dembélé, Moctar, and Sander J. Zwart. 2016. Evaluation and Comparison of Satellite-Based
Rainfall Products in Burkina Faso, West Africa. International Journal of Remote Sensing 37
(17): 39954014. https://doi.org/10.1080/01431161.2016.1207258.
Derin, Yagmur, and Koray K. Yilmaz. 2014. Evaluation of Multiple Satellite-Based Precipitation
Products over Complex Topography. Journal of Hydrometeorology 15 (4): 14981516.
https://doi.org/10.1175/JHM-D-13-0191.1.
Funk, Chris, Pete Peterson, Martin Landsfeld, Diego Pedreros, James Verdin, Shraddhanand
Shukla, Gregory Husak, et al. 2015. The Climate Hazards Infrared Precipitation with Stations
- A New Environmental Record for Monitoring Extremes. Scientific Data.
https://doi.org/10.1038/sdata.2015.66.
Gao, Y. C., and M. F. Liu. 2013. Evaluation of High-Resolution Satellite Precipitation Products Using
Rain Gauge Observations over the Tibetan Plateau. Hydrology and Earth System Sciences 17
(2): 83749. https://doi.org/10.5194/hess-17-837-2013.
Gosset, Marielle, Julien Viarre, Guillaume Quantin, and Matias Alcoba. 2013. Evaluation of Several
Rainfall Products Used for Hydrological Applications over West Africa Using Two High-
Resolution Gauge Networks. Quarterly Journal of the Royal Meteorological Society 139 (673):
92340. https://doi.org/10.1002/qj.2130.
Hersbach, Hans, Bill Bell, Paul Berrisford, Shoji Hirahara, András Horányi, Joaquín Muñoz-Sabater,
Julien Nicolas, et al. 2020. The ERA5 Global Reanalysis. Quarterly Journal of the Royal
Meteorological Society. https://doi.org/10.1002/qj.3803.
Lebel, Thierry, Arona Diedhiou, and Henri Laurent. 2003. Seasonal Cycle and Interannual Variability
of the Sahelian Rainfall at Hydrological Scales. Journal of Geophysical Research:
Atmospheres 108 (8): 111. https://doi.org/10.1029/2001jd001580.
Nicholson, Sharon E., B. Some, J. McCollum, E. Nelkin, D. Klotter, Y. Berte, B. M. Diallo, et al.
2003a. Validation of TRMM and Other Rainfall Estimates with a High-Density Gauge Dataset
for West Africa. Part I: Validation of GPCC Rainfall Product and Pre-TRMM Satellite and
Blended Products. Journal of Applied Meteorology 42 (10): 133754.
https://doi.org/10.1175/1520-0450(2003)042<1337:VOTAOR>2.0.CO;2.
———. 2003b. Validation of TRMM and Other Rainfall Estimates with a High-Density Gauge
Dataset for West Africa. Part II: Validation of TRMM Rainfall Products Blended Products.
Journal of Applied Meteorology 42 (10): 133754. https://doi.org/10.1175/1520-
0450(2003)042<1337:VOTAOR>2.0.CO;2.
Nicholson, Sharon E. 2013. The West African Sahel: A Review of Recent Studies on the Rainfall
Regime and Its Interannual Variability 2013.
Panthou, G., T. Lebel, T. Vischel, G. Quantin, Y. Sane, A. Ba, O. Ndiaye, A. Diongue-Niang, and M.
16
Diopkane. 2018. Rainfall Intensification in Tropical Semi-Arid Regions: The Sahelian Case.
Environmental Research Letters. https://doi.org/10.1088/1748-9326/aac334.
Pierre, Caroline, Gilles Bergametti, Béatrice Marticorena, Eric Mougin, Thierry Lebel, and Abdou Ali.
2011. Pluriannual Comparisons of Satellite-Based Rainfall Products over the Sahelian Belt for
Seasonal Vegetation Modeling. Journal of Geophysical Research Atmospheres 116 (18): 1
25. https://doi.org/10.1029/2011JD016115.
Satgé, Frédéric, Dimitri Defrance, Benjamin Sultan, Marie Paule Bonnet, Frédérique Seyler, Nathalie
Rouché, Fabrice Pierron, and Jean Emmanuel Paturel. 2020. Evaluation of 23 Gridded
Precipitation Datasets across West Africa. Journal of Hydrology 581 (November 2019).
https://doi.org/10.1016/j.jhydrol.2019.124412.
Sultan, B., P. Roudier, P. Quirion, A. Alhassane, B. Muller, M. Dingkuhn, P. Ciais, M. Guimberteau,
S. Traore, and C. Baron. 2013. Assessing Climate Change Impacts on Sorghum and Millet
Yields in the Sudanian and Sahelian Savannas of West Africa. Environmental Research
Letters 8 (1). https://doi.org/10.1088/1748-9326/8/1/014040.
Sultan, Benjamin, Bruno Barbier, Jeanne Fortilus, Serigne Modou Mbaye, and Grégoire Leclerc.
2010. Estimating the Potential Economic Value of Seasonal Forecasts in West Africa: A Long-
Term Ex-Ante Assessment in Senegal. Weather, Climate, and Society 2 (1): 6987.
https://doi.org/10.1175/2009WCAS1022.1.
Tang, Guoqiang, Martyn P. Clark, Simon Michael Papalexiou, Ziqiang Ma, and Yang Hong. 2020.
Have Satellite Precipitation Products Improved over Last Two Decades? A Comprehensive
Comparison of GPM IMERG with Nine Satellite and Reanalysis Datasets. Remote Sensing of
Environment 240 (January): 111697. https://doi.org/10.1016/j.rse.2020.111697.
USGS. 2016. West Africa: Land Use and Land Cover Dynamics. U.S. Geological Survey - Earth
Resources Observation and Science (EROS) Center.
https://eros.usgs.gov/westafrica/node/147.
17
Appendix
Seasonal correlation is calculated based on the seasonal precipitation of each region in 35 years
(i.e. n = 35).
Figure 9. Correlation of seasonal precipitation of ERA5 and BADOPLU.
Figure 10. Correlation of seasonal precipitation of CHIRS and BADOPLU.
18
Figure 11. Number of stations per year (provided by Guillaume Quantin 2020).
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Within the Copernicus Climate Change Service (C3S), ECMWF is producing the ERA5 reanalysis, which embodies a detailed record of the global atmosphere, land surface and ocean waves from 1950 onwards once completed. This new reanalysis replaces the ERA‐Interim reanalysis that was started in 2006 (spanning 1979 onwards). ERA5 is based on the Integrated Forecasting System (IFS) Cy41r2 which was operational in 2016. ERA5 thus benefits from a decade of developments in model physics, core dynamics and data assimilation. In addition to a significantly enhanced horizontal resolution of 31 km, compared to 80 km for ERA‐Interim, ERA5 has hourly output throughout, and an uncertainty estimate from an ensemble (3‐hourly at half the horizontal resolution). This paper describes the general setup of ERA5, as well as a basic evaluation of characteristics and performance. Focus is on the dataset from 1979 onwards that is currently publicly available. Re‐forecasts from ERA5 analyses show a gain of up to one day in skill with respect to ERA‐Interim. Comparison with radiosonde and PILOT data prior to assimilation shows an improved fit for temperature, wind and humidity in the troposphere, but not the stratosphere. A comparison with independent buoy data shows a much improved fit for ocean wave height. The uncertainty estimate reflects the evolution of the observing systems used in ERA5. The enhanced temporal and spatial resolution allows for a detailed evolution of weather systems. For precipitation, global‐mean correlation with monthly‐mean GPCP data is increased from 67% to 77%. In general low‐frequency variability is found to be well‐represented and from 10 hPa downwards general patterns of anomalies in temperature match those from the ERA‐Interim, MERRA‐2 and JRA‐55 reanalyses. This article is protected by copyright. All rights reserved.
Article
Full-text available
The rainfall regime of West Africa is highly variable over a large range of space and time scales. With rainfall agriculture being predominent in the region, the local population is extremely vulnerable to intraseasonal dry spells and multi-year droughts as well as to intense rainfall over small time steps. Were this variability to increase, it might render the area close from becoming unhabitable. Anticipating any change is thus crucial from both a societal and a scientific perspective. Despite continuous efforts in Global Climate Model (GCM) development, there is still no agreement on the sign of the future rainfall regime change in the region. Regional Climate Models (RCMs) are used for more accurate projections of future changes as well as end-user-oriented impact studies. In this study, the sensitivity of the Modèle Atmosphérique Régional (MAR) to homogeneous perturbations in boundary forcing air temperature and/or SST is assessed with the aim to better understand (i) the thermodynamical imprint of the recent rainfall regime changes and (ii) the impact of errors in driving data on the West African rainfall regime simulated by an RCM. After an evaluation step where the model is proved to satisfactorily simulate the West African Monsoon (WAM), sensitivity experiments display contrasted, sizable and robust responses of the simulated rainfall regime. The rainfall responses to the boundary forcing perturbations compare in magnitude with the intrinsic model bias, giving support for such an analysis. A physical interpretation of the rainfall anomalies provides confidence in the model response consistency and shows the potential of such an experimental protocol for future climate change downscalling over this region.
Article
Full-text available
An anticipated consequence of ongoing global warming is the intensification of the rainfall regimes meaning longer dry spells and heavier precipitation when it rains, with potentially high hydrological and socio-economic impacts. The semi-arid regions of the intertropical band, such as the Sahel, are facing particularly serious challenges in this respect since their population is strongly vulnerable to extreme climatic events. Detecting long term trends in the Sahelian rainfall regime is thus of great societal importance, while being scientifically challenging because datasets allowing for such detection studies are rare in this region. This study addresses this challenge by making use of a large set of daily rain gauge data covering the Sahel (defined in this study as extending from 20°W–10°E and from 11°N–18°N) since 1950, combined with an unparalleled 5 minute rainfall observations available since 1990 over the AMMA-CATCH Niger observatory. The analysis of the daily data leads to the assertion that a hydro-climatic intensification is actually taking place in the Sahel, with an increasing mean intensity of rainy days associated with a higher frequency of heavy rainfall. This leads in turn to highlight that the return to wetter annual rainfall conditions since the beginning of the 2000s—succeeding the 1970–2000 drought—is by no mean a recovery towards the much smoother regime that prevailed during the 1950s and 1960s. It also provides a vision of the contrasts existing between the West Sahel and the East Sahel, the East Sahel experiencing a stronger increase of extreme rainfall. This regional vision is complemented by a local study at sub-daily timescales carried out thanks to the 5 minute rainfall series of the AMMA-CATCH Niger observatory (12000 km²). The increasing intensity of extreme rainfall is also visible at sub-daily timescales, the annual maximum intensities have increased at an average rate of 2%–6% per decade since 1990 for timescales ranging from 5 min to 1 hour. Both visions—regional/long term/daily on the one hand, and local/27/years/sub-daily, on the other—converge to the conclusion that, rather than a rainfall recovery, the Sahel is experiencing a new era of climate extremes that roughly started at the beginning of this century.
Article
Full-text available
The paper aimed at assessing the capabilities and limitations of five different precipitation products to describe rainfall over West Africa. Five gridded precipitation datasets of the Tropical Rainfall Measurement Mission (TRMM) Multi-Platform Analysis (TMPA 3B43v7); University of Delaware (UDEL version 3.01); Climatic Research Unit (CRU version 3.1); Global Precipitation Climatology Centre (GPCC version 7) and African Rainfall Climatology (ARC version 2) were compared and validated with reference ground observation data from 81 stations spanning a 19-year period, from January 1990 to December 2008. Spatial investigation of the precipitation datasets was performed, and their capability to replicate the inter-annual and intra-seasonal variability was also assessed. The ability of the products to capture the El Nino and La Nina events were also assessed. Results show that all the five datasets depicted similar spatial distribution of mean rainfall climatology, although differences exist in the total rainfall amount for each precipitation dataset. Further analysis shows that the three distinct phases of the mean annual cycle of the West Africa Monsoon precipitation were well captured by the datasets. However, CRU, GPCC and UDEL failed to capture the little dry season in the month of August while UDEL and GPCC underestimated rainfall amount in the Sahel region. Results of the inter-annual precipitation anomalies shows that ARC2 fail to capture about 46% of the observed variability while the other four datasets exhibits a greater performance (r > 0.9). All the precipitation dataset except ARC2 were consistent with the ground observation in capturing the dry and wet conditions associated with El Nino and La Nina events, respectively. ARC2 tends to overestimate the El Nino event and failed to capture the La Nina event in all the years considered. In general GPCC, CRU and TRMM were found to be the most outstanding datasets and can, therefore, be used for precipitation assessment over West Africa.
Article
Full-text available
High-resolution satellite precipitation products are very attractive for studying the hydrologic processes in mountainous areas where rain gauges are generally sparse. Four high-resolution satellite precipitation products are evaluated using gauge measurements over different climate zones of the Tibetan Plateau (TP) within a 6 yr period from 2004 to 2009. The four satellite-based precipitation data sets are: Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis 3B42 version 6 (TMPA) and its Real Time version (TMPART), Climate Prediction Center Morphing Technique (CMOPRH) and Precipitation Estimation from Remotely Sensed Information using Artificial Neural Network (PERSIANN). TMPA and CMORPH, with higher correlation coefficients and lower root mean square errors (RMSEs), show overall better performance than PERSIANN and TMPART. TMPA has the lowest biases among the four precipitation data sets, which is likely due to the correction process against the monthly gauge observations from global precipitation climatology project (GPCP). TMPA also shows large improvement over TMPART, indicating the importance of gauge-based correction on accuracy of rainfall. The four products show better agreement with gauge measurements over humid regions than that over arid regions where correlation coefficients are less than 0.5. Moreover, the four precipitation products generally tend to overestimate light rainfall (0–10 mm) and underestimate moderate and heavy rainfall (>10 mm). Moreover, this study extracts 24 topographic variables from a DEM (digital elevation model) and uses a linear regression model to explore the bias–topography relationship. Results show that biases of TMPA and CMORPH present weak dependence on topography. However, biases of TMPART and PERSIANN present dependence on topography and variability of elevation and surface roughness plays important roles in explaining their biases.
Article
Full-text available
The performance of seven operational high-resolution satellite-based rainfall products – Africa Rainfall Estimate Climatology (ARC 2.0), Climate Hazards Group InfraRed Precipitation with Stations (CHIRPS), Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN), African Rainfall Estimation (RFE 2.0), Tropical Applications of Meteorology using SATellite (TAMSAT), African Rainfall Climatology and Time-series (TARCAT), and Tropical Rainfall Measuring Mission (TRMM) daily and monthly estimates – was investigated for Burkina Faso. These were compared to ground data for 2001–2014 on a point-to-pixel basis at daily to annual time steps. Continuous statistics was used to assess their performance in estimating and reproducing rainfall amounts, and categorical statistics to evaluate rain detection capabilities. The north–south gradient of rainfall was captured by all products, which generally detected heavy rainfall events, but showed low correlation for rainfall amounts. At daily scale they performed poorly. As the time step increased, the performance improved. All (except TARCAT) provided excellent scores for Bias and Nash–Sutcliffe Efficiency coefficients, and overestimated rainfall amounts at the annual scale. RFE performed the best, whereas TARCAT was the weakest. Choice of product depends on the specific application: ARC, RFE, and TARCAT for drought monitoring, and PERSIANN, CHIRPS, and TRMM daily for flood monitoring in Burkina Faso. © 2016 Africa Rice Centre. Published by Informa UK Limited, trading as Taylor & Francis Group.
Article
Full-text available
The Climate Hazards group Infrared Precipitation with Stations (CHIRPS) dataset builds on previous approaches to ‘smart’ interpolation techniques and high resolution, long period of record precipitation estimates based on infrared Cold Cloud Duration (CCD) observations. The algorithm i) is built around a 0.05° climatology that incorporates satellite information to represent sparsely gauged locations, ii) incorporates daily, pentadal, and monthly 1981-present 0.05° CCD-based precipitation estimates, iii) blends station data to produce a preliminary information product with a latency of about 2 days and a final product with an average latency of about 3 weeks, and iv) uses a novel blending procedure incorporating the spatial correlation structure of CCD-estimates to assign interpolation weights. We present the CHIRPS algorithm, global and regional validation results, and show how CHIRPS can be used to quantify the hydrologic impacts of decreasing precipitation and rising air temperatures in the Greater Horn of Africa. Using the Variable Infiltration Capacity model, we show that CHIRPS can support effective hydrologic forecasts and trend analyses in southeastern Ethiopia.
Article
The Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG) produces the latest generation of satellite precipitation estimates and has been widely used since its release in 2014. IMERG V06 provides global rainfall and snowfall data beginning from 2000. This study comprehensively analyzes the quality of the IMERG product at daily and hourly scales in China from 2000 to 2018 with special attention paid to snowfall estimates. The performance of IMERG is compared with nine satellite and reanalysis products (TRMM 3B42, CMORPH, PERSIANN-CDR, GSMaP, CHIRPS, SM2RAIN, ERA5, ERA-Interim, and MERRA2). Results show that the IMERG product outperforms other datasets, except the Global Satellite Mapping of Precipitation (GSMaP), which uses daily-scale station data to adjust satellite precipitation estimates. The monthly-scale station data adjustment used by IMERG naturally has a limited impact on estimates of precipitation occurrence and intensity at the daily and hourly time scales. The quality of IMERG has improved over time, attributed to the increasing number of passive microwave samples. SM2RAIN, ERA5, and MERRA2 also exhibit increasing accuracy with time that may cause variable performance in climatological studies. Even relying on monthly station data adjustments, IMERG shows good performance in both accuracy metrics at hourly time scales and the representation of diurnal cycles. In contrast, although ERA5 is acceptable at the daily scale, it degrades at the hourly scale due to the limitation in reproducing the peak time, magnitude and variation of diurnal cycles. IMERG underestimates snowfall compared with gauge and reanalysis data. The triple collocation analysis suggests that IMERG snowfall is worse than reanalysis and gauge data, which partly results in the degraded quality of IMERG in cold climates. This study demonstrates new findings on the uncertainties of various precipitation products and identifies potential directions for algorithm improvement. The results of this study will be useful for both developers and users of satellite rainfall products.
Article
This study aims reporting on 23 gridded precipitation datasets (P-datasets) reliability across West Africa through direct comparisons with rain gauges measurement at the daily and monthly time scales over a 4 years period (2000-2003). All P-datasets reliability vary in space and time. The most efficient P-dataset in term of Kling–Gupta Efficiency (KGE) changes at the local scale and the P-dataset performance is sensitive to seasonal effects. Satellite-based P-datasets performed better during the wet than the dry season whereas the opposite is observed for reanalysis P-datasets. The best overall performance was obtained for MSWEP v.2.2 and CHIRPS v.2 for daily and monthly time-step, respectively. Part of the differences in P-dataset performance at daily and monthly time step comes from the time step used to proceed the gauges adjustment (i.e day or month) and from a mismatch between gauge and satellite reporting times. In comparison to the others P-datasets, TMPA-Adj v.7 reliability is stable and reach the second highest KGE value at both daily and monthly time step. Reanalysis P-datasets (WFDEI, MERRA-2, JRA-55, ERA-Interim) present among the lowest statistical scores at the daily time step, which drastically increased at the monthly time step for WFDEI and MERRA-2. The non-adjusted P-datasets were the less efficient, but, their near-real time availability should be helpful for risk forecast studies (i.e. GSMaP-RT v.6). The results of this study give important elements to select the most adapted P-dataset for specific application across West Africa.
Article
The evaluation of rainfall products over the West African region will be an important component of the Megha-Tropiques (MT) Ground Validation (GV) plan. In this paper, two dense research gauge networks from Benin and Niger, integrated in the MT GV plan, are presented and are used to evaluate several currently available global or regional satellite-based rainfall products. Eight products—the Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN), Climate Prediction Center Morphing method (CMORPH), Tropical Rainfall Measuring Mission (TRMM) Multi-satellite Precipitation Analysis (TMPA) 3B42 real-time and gauge-adjusted version, Global Satellite Mapping of Precipitation (GSMaP), Climate Prediction Center (CPC) African Rainfall Estimate (RFE), Estimation des Precipitation par SATellite (EPSAT), and Global Precipitation Climatology Project One Degree Daily estimate (GPCP-1DD)—are compared to the ground reference. The comparisons are carried out at daily, 1° resolution, over the rainy season (June–September), between the years 2003 and 2010. The work focuses on the ability of the various products to reproduce salient features of the rainfall regime that impact the hydrological response. The products are analysed on a multi-criteria basis, focusing in particular on the way they distribute the rainfall within the season and by rain rate class. Standard statistical diagnoses such as the correlation coefficient, bias, root mean square error and Nash skill score are computed and the inter-annual variability is documented. Two simplified hydrological models are used to illustrate how the nature and structure of the product error impact the model output in terms of runoff (calculated using the Soil Conservation Service method, SCS, in Niger) or outflow (calculated with the ‘modèle du Génie Rural à 4 paramètres Journalier’, GR4J model, in Benin). Copyright © 2013 Royal Meteorological Society