Coffee Yield and Production : Rwanda

EARS Coffee Yield & Production Estimate




RWANDA - 15 October 2008


Subject:
Coffee Yield and Production
Country:
Rwanda
Issuing date: 15 October 2008


INTRODUCTION

The present document provides an estimate of the coffee yield and production for 2008 in
Rwanda. The results are based on historical satellite data on drought and historical data on
coffee production and were extrapolated to the drought situation of 2007-2008.


METHOD
The assessment of coffee growing conditions and the coffee yield and production estimates are
based on visible and thermal infrared hourly data. The coffee yield estimates for Rwanda were
produced as following:
(1) Hourly Meteosat data are processed to daily average values of surface temperature, air
temperature, global radiation, net radiation, potential and actual evapotranspiration.
(2) Radiation and evapotranspiration data are related to data on coffee yield and coffee cultivation
areas in Rwanda. The coffee yield response to water is quantified and a spatially distributed map
of coffee yield based on the satellite data of 2008 is produced.
(3) The distributed yield results are integrated for coffee growing areas, country and provinces.
Urban areas, forest, water and barren land are excluded.


COFFEE
The crop calendar in Figure 1 shows the vegetative period (green), the flowering and yield
formation period (grey) and the harvesting period (yellow) for coffee in Rwanda.

Jan Feb Mar Apr May Jun Jul
Aug Sep Oct Nov Dec
RWANDA



Figure 1: Coffee Growing Calendar Rwanda


The yield response of crops to drought and water deficits during the growing season is usually
quantified by the yield response factor Ky, based on the relationship described by Doorenbos &
Kassam (1986).
For coffee however, the Ky yield response factor is not known and surprisingly little has been
reported on yield responses to drought. For this reason, EARS has established the relationship
between drought and coffee yield for Rwanda using historical satellite data and historical data on
coffee production. A relationship between the satellite-based agroclimatological data and reported
coffee yields is built and used to provide estimates for 2008 production.
EARS – Satellite Data for Water and Food
1

---

EARS Coffee Yield & Production Estimate




RWANDA - 15 October 2008


DROUGHT

Coffee is most sensitive during the flowering and bean formation period which lasts from October
until March in Rwanda. The maps below in figure 2 show the average relative evapotranspiration
maps for the flowering and yield formation season from October until March for each of the five
historical years investigated (2002-2007). Relative evapotranspiration is derived from remote
sensing data and is an agricultural drought indicator. Lower values (red) represent dryer
conditions.

.

Figure 2: Relative Evapotranspiration (%) from October until March for five years (2002-2007)

Relative evapotranspiration is defined as the ratio of actual to potential evapotranspiration
(LE/LEp). Actual evapotranspiration (LE) represents the real amount of water that is lost from the
earth surface and the vegetation into the atmosphere. Potential evapotranspiration (LEp) is the
amount of water that could be evaporated and transpired if there was sufficient water available.
Actual evapotranspiration can never be greater than potential evapotranspiration, but can be
lower if there is not enough water to be evaporated or plants are unable to readily transpire.
Relative evapotranspiration hence quantifies the amount of water that is available relative to the
amount of water that is demanded by the atmosphere and is the most useful parameter to
indicate the severity of agricultural drought.
EARS – Satellite Data for Water and Food
2

---

EARS Coffee Yield & Production Estimate




RWANDA - 15 October 2008


GROUND DATA

The historical data of national coffee production (tonnes) from the Coffee Board of Rwanda and
coffee cultivation surface areas from FAOstat were used for each of the five historical years 2003-
2007 to build the relationship. The data are shown in following table:

Table 1: Historical coffee production and yields
Cultivated
Production
Yield

Area

(tonnes)
(ha)
(kg/ha)
2003
14726
20000
736
2004
27027
29000
932
2005
17500
27000
648
2006
26876
25000
1075
2007
14264
24000
594


DROUGHT AND YIELD RESPONSE

The relationship between coffee yield (kg/ha) and relative evapotranspiration that was established
for Rwanda is shown in Figure 3. The relative evapotranspiration was averaged over the more
dense coffee growing areas in Rwanda (> 5000 trees/km2). The data point for 2006 was omitted
because the high production can not be explained by the drought situation, 2005-2006 being the
driest season of all five years.

1000
900
800
)
a
700
/
h
g
600

(
k
l
d
i
e
500

Y
e
400
f
f
e
o
y = 39.023x - 1805.8
C
300
R2 = 0.9031
200
100
0
50
55
60
65
70
75
Relative Evapotranspiration (% )

Figure 3: Coffee yield response to drought 2002-2007, Rwanda

The relationship shows a strong correlation with a regression coefficient of 0.9. Coffee yield
increases by 39 kg/ha with 1% increase in relative evapotranspiration and an offset of 1806
kg/ha.

Coffee Yield =39 x Relative Evapotranspiration – 1806

EARS – Satellite Data for Water and Food
3

---

EARS Coffee Yield & Production Estimate




RWANDA - 15 October 2008


MAPS

The maps on the next page show the Meteosat relative evapotranspiration (figure 4) and the final
coffee yield production estimate (figure 5) for 2008. The coffee growing areas and coffee tree
density is shown in figure 6. The final results were integrated over the more dense coffee growing
areas (pink and red).


Figure 4: Relative Evapotranspiration (%) from October 2007 to March 2008


Figure 5: Coffee Yield (kg/ha) estimate for 2008
EARS – Satellite Data for Water and Food
4

---

EARS Coffee Yield & Production Estimate




RWANDA - 15 October 2008


Figure 6: Coffee growing areas ©CGIS-NUR
EARS – Satellite Data for Water and Food
5

---

EARS Coffee Yield & Production Estimate




RWANDA - 15 October 2008


COFFEE YIELD AND PRODUCTION ESTIMATES FOR RWANDA 2008

With the relative evapotranspiration data and the yield response relationship, estimates of yields
for 2008 are provided by province, averaging over the main coffee growing areas. The surface
area of coffee cultivation reported in 2007 is used to provide the production estimates. The total
surface area is proportionally attributed to the different provinces.

The following tabulated estimates by province are provided:

•
Coffee yields (kg/ha)
•
Coffee cultivation area (ha)
•
Coffee production (tonnes)

RWANDA
Surface
Yield
Production

Area

(kg/ha)
(ha)
(tonnes)
BYUMBA
810
2405
1948
GISENYI
1100
740
814
KIGALI-NGALI
620
5457
3383
GITARAMA
630
4254
2680
KIBUNGO
630
4855
3059
KIBUYE
870
879
764
CYANGUGU
1200
832
999
GIKONGORO
710
925
657
BUTARE
580
3653
2119
TOTAL
794
24000
16423


REFERENCE
Doorenbos, J., Kassam, A. H. (1986). Yield response to water. FAO irrigation and drainage paper
33. Food and Agricultural Organization of the United Nations, Rome.



ADMINISTRATIVE BOUNDARIES
National, regional or provincial averages are calculated based on the administrative boundary
layers from the Global Administrative Unit Layers (GAUL) system.

CONTACT

For further information please contact:

EARS Earth Environment Monitoring BV
Kanaalweg 1, 2628 EB DELFT, the Netherlands
Telephone: +31-15-2562404, Fax: +31-15-2623857
Email: marjolein@ears.nl

Website: www.ears.nl
EARS – Satellite Data for Water and Food
6

---