Advertisement

Earth Systems and Environment

, Volume 2, Issue 2, pp 281–292 | Cite as

Modeling the Spatio-Temporal Meteorological Drought Characteristics Using the Standardized Precipitation Index (SPI) in Raya and Its Environs, Northern Ethiopia

  • Eskinder Gidey
  • Oagile Dikinya
  • Reuben Sebego
  • Eagilwe Segosebe
  • Amanuel Zenebe
Original Article

Abstract

In Ethiopia, recurrent droughts were associated with El Niño phenomenon, particularly in the study area where the majority of smallholder farmers depend on rain-fed agriculture. The aim of this study was to model the spatio-temporal meteorological drought onset, offset, duration, magnitude, intensity, frequency, severity and spatial extent in Raya and its environs, Northern Ethiopia. Both the ground meteorological stations and TAMSAT (tropical applications of meteorology using satellite and ground-based observations) monthly rainfall data were gathered from the National Meteorological Agency of Ethiopia for the period 1983–2015. The Inverse Distance Weighting (IDW), a type of the deterministic interpolation method was applied to quantify the amount of seasonal rainfall by producing surface rainfall map. A Standardized Precipitation Index (SPI) at 3-month timescale was used to evaluate seasonal rainfall deficit and characterize meteorological drought. The results revealed that all sites obtained minimal and irregular rainfall, hence led to catastrophic droughts. Also, the findings showed that there was high rainfall variability across the study area that ranged from 28.14 to 42.32%. As a result, mild-to-severe meteorological drought phenomena were observed once in every 2–3 years. This incidence was found to be high in terms of spatial and temporal coverage during the last three decades. Therefore, this study may help to offer better insight for policymakers to establish drought mitigation and adaptation strategies.

Keywords

Meteorological drought characteristics Rainfall SPI-3 Raya Ethiopia 

1 Introduction

Drought has been caused by the decline of seasonal or annual rainfall in a long period (McKee et al. 1993; Peters et al. 2002). Wilhite (1996) reported that the extended periods of acute water shortage due to meteorological drought are increasing at an alarming rate in both developed and developing countries. Belal et al. (2014) stated that one-third of the world’s population resides in water scarce areas, and nearly 1.1 billion people lack access to clean drinking water. Studies showed that people dwelling in areas with high water stresses projected to increase significantly from 300 million to 3 billion by the year 2025. In developing countries, e.g., Ethiopia, the impacts of drought are more devastating, leading to food shortages, population migration, and livestock mortality (Kogan et al. 2013). Abebe (2014 ) and Tabari et al. (2015) reported that water is a scarce resource in the northern highlands of Ethiopia due to high rainfall variability and recurrent drought. Besides, the area is characterized by dry or arid and semi-arid climatic conditions. As a result, the incidence of drought over the region has spatial and temporal variations and it becomes critical to monitor the relative drought occurrence levels and their distributions over a region at a time (Wang et al. 2001). Şen et al. (2017) reported that arid regions experience both high temperature and evapotranspiration. That is why drought occurred repeatedly in all agro-climatic zones of the study area such as lowlands, midlands, and highlands in the last three decades. World Bank (2006) reported that the challenge has gradually expanded from the north to the rest of the country and its intensity has deepened. However, its onset, offset, spatial extent, magnitude, intensity, duration, frequency, and severity differ both in spatial and temporal coverage. Svoboda et al. (2012) reported that this phenomenon might extend over a season or longer period, leading to inadequate rainfall to meet the demands of humans, vegetation and livestock. This may cause complex environmental issues and impacts, particularly on the economy and society (Shamsipour et al. 2011). For instance, in the year 2015, the study area experienced its worst drought because of El Niño resulting in failure of the main rains (July, August, and September) season in the most parts. The failure of two successive main rainy seasons, which feeds nearly 80–85% of the population, caused an increase in malnutrition rates due to food shortage and a high shortage of forages for livestock. World Bank (2006) indicated that food shortages during drought period drive people to strip natural resources (e.g., fuelwood, craft supplies, and wild foods). This led to a significant effect on the livelihood of the community that relied on rain-fed agriculture as well as the regional or national economic growth at large.

The incidence of meteorological drought has largely escalated because of the extended periods of rainfall deficiency (Wilhite 1996). Gidey et al. (2017) reported that the long-term rainfall condition of the study area was up to 558 mm and its distribution was irregular. As a result, the manifestations of meteorological drought became severe. Therefore, this natural hazard needs to be monitored scientifically to mitigate its negative impacts. Tabari et al. (2015) stated that there is a need to understand the previous drought events using a scientific method of assessing the condition rainfall variability in Ethiopia. Gidey et al. (2018) reported that several meteorological drought indices have been developed to measure the incidence of drought. Conversely, the Standardized Precipitation Index (SPI) was found to be important to evaluate the extent of rainfall deficit in a given period for two reasons: (1) it needs only rainfall data; (2) it can be applied worldwide on different timescales such as 1, 3, 6, 9, 12, and 24 months. The index also provides early warning information to formulate appropriate mitigation measures locally, regionally and contently. This may support policy improvement, decision-making and build up a drought resilience strategic framework in the economic development of the community. For example, selection of an appropriate timescale helps in in-depth assessment of the nature of drought such as frequency, magnitude, and duration (McKee et al. 1993). Studying of meteorological drought over a short timescale would have great significance to small and marginal farmers as well as decision-makers to introduce proper water resource management mechanisms in areas severely affected by drought like Raya and its environs, Ethiopia. The aim of this study was to model the spatio-temporal meteorological drought onset, offset, duration, magnitude, intensity, severity, frequency, and extent using the Standardized Precipitation Index (SPI) at 3-month timescale in Raya and its environs, Northern Ethiopia.

2 Materials and Methods

2.1 Study Area

The study area is found at 39°24′40′′ and 40°25′20′′ Easting and 12°7′20′′ and 13°8′0′′ Northing (Fig. 1) (Gidey et al. 2017). The area has 11 districts located in three agro-ecological zones such as lowland (e.g., Gulina, Yalo, Megale), midland (e.g., Alamata, Kobo, Hintalo Wejirat, Raya Azebo), and highlands (e.g., Gidan, Alaje, Ofla, and Endamehoni). Gidey et al. (2017) reported that the study area covers about 14,532 km2. Ayenew et al. (2013) and Gidey et al. (2017) indicated that the rainfall status of the area is irregular and it receives up to 558 mm per annum. Gidey et al. (2017) stated that the maximum temperature (Tmax) and minimum temperature (Tmin) of the study area was about 30.5 and 15.9 °C, respectively, in 2015. During this period, the highest temperature was observed during the last three decades (Gidey et al. 2017). The maximum and minimum elevation value of the study area is 4129 and 324 m above sea level (m.a.s.l) (Fig. 1). The mean elevation value of the area is also 1762 m above sea level (m.a.s.l) (Gidey et al. 2017). Eutric cambisols is the major soil type and covers about 4667.1 km2 (32.1%), while the dystric gleysols cover only small portions of the area, i.e., about 1.1 km2 or 0.001%, respectively (Gidey et al. 2017).
Fig. 1

Map of the study area

(improved: Gidey et al. 2017)

2.2 Data Collection

2.2.1 Rainfall

Rainfall data are the most important inputs for assessing meteorological drought characteristics (Gidey et al. 2018). Table 1 shows the total number of meteorological stations used to collect the monthly rainfall data from the period 1983–2015. However, the rainfall data had some missing values due to malfunctioning of the recording instruments and other related problems. To resolve this challenge, the tropical applications of meteorology using satellite and ground-based observations (TAMSAT) monthly rainfall data at 4 km by 4 km grid size were collected from the National Meteorological Agency (NMA) of Ethiopia (Gidey et al. 2018). In this case, the satellite data were used as a proxy for observational gaps in weather stations (Kogan et al. 2013; Kogan and Guo 2016).
Table 1

Spatial distribution of meteorological stations found in the study area

(soure: Gidey et al. 2018)

S. no

Station

Location

Longitude (E)

Latitude (N)

Altitude (m.a.s.l)

1.

Adigudom

39.51

13.25

2100

2.

Adishehu

39.53

12.93

2465

3.

Alamata

39.71

12.42

1589

4.

Chercher

39.77

12.54

1781

5.

Debub

39.64

13.08

2536

6.

Erebti

39.99

13.25

877

7.

Hashenge

39.52

12.60

2480

8.

Kelawn

39.98

12.14

859

9.

Kobo

39.63

12.13

1470

10.

Korem

39.51

12.51

2450

11.

Mehoni

39.65

12.80

1590

12.

Maichew

39.53

12.78

2432

13.

Robet

39.62

12.01

1629

14.

Sanka

39.48

11.89

1989

15.

Tekulish

39.48

12.15

2181

16.

Waja

39.60

12.30

1458

17.

Yalo

39.88

12.36

869

18.

Zata

39.28

12.51

2334

19.

Zobel

39.75

12.17

1893

2.3 Data Processing and Analysis

2.3.1 Rainfall

The Inverse Distance Weighting (IDW) a type of the deterministic interpolation method was applied to analyze the seasonal of rainfall and produce a surface rainfall map by predicting the values of unmeasured stations based on the known gauge station values using ArcGIS 10.4.1 geostatistical tools (Gidey et al. 2018). This interpolation technique is widely accepted because of its flexibility and reliability. That is why various studies have reported that IDW is better if the gauge stations are well distributed. In this study, the interpolated rainfall data were used as input to assess the values of SPI in each district and determine the historical meteorological drought onset, offset, duration, magnitude, intensity, severity, frequency and spatial extent.

2.3.2 Assessment of Rainfall Variability

In this study, the coefficient of variation (CV) was applied to assess the rainfall variability as follows (Eqs. 1, 2):
$$ \text{CV} = \sigma /\bar{x}, $$
(1)
$$ \text{CV} (\% ) = 100 \times \frac{\sigma }{{\bar{x}}}, $$
(2)
where CV(%) = coefficient of variation in  %, σ = standard deviation, and \( \bar{x} \) = mean of rainfall (1983–2015).

2.3.3 Meteorological Drought Monitoring Using SPI

In 2009, the World Meteorological Organization (WMO) has accepted the Standardized Precipitation Index (SPI) to be a universal meteorological drought index. It also emphasized that all National Meteorological and Hydrological Services found around the world should apply SPI to monitor meteorological droughts. Studies have indicated that SPI is applicable at any location based on the long-term rainfall data. The index is helpful for assessing the wetness and dryness conditions. In this study, the SPI algorithm developed by McKee et al. (1993) was applied to monitor meteorological drought. The index was evaluated by transforming the cumulative probability of rainfall data into the standard normal distribution (AghaKouchak et al. 2015; Rossi et al. 2007; McKee et al. 1993). McKee et al. (1993) reported that SPI values range between − 1 and 1. The negative SPI value shows periods of drought stress, while the positive value shows non-drought conditions. Mishra and Singh (2010) revealed that the strength of SPI is capability to quantify for different timescales. However, Bonaccorso et al. (2003) suggested that SPI on a short timescale like 3 or 6 months is appropriate for detecting drought events that affect agricultural practices, while the longer timescale, such as 12 or 24 months is more suitable for water resources management purposes. In this study, the SPI was calculated for 3 month timescale after the monthly rainfall data were arranged (Fig. 2). Studies showed that the long-term rainfall data are not normally distributed. As a result, it requires the estimation of gamma distribution parameters such as α and β, which would then transform into a normal distribution as follows (Eq. 3):
$$ g(x) = \tfrac{1}{{\beta^{\alpha } \varGamma (\alpha )}}x^{\alpha - 1} {\text{e}}^{ - x/\beta } \quad \;{\text{for}}\;x > 0, $$
(3)
where α = shape parameter, β =  scale factor, x =  the amount of rainfall, and Γ = gamma function.
Fig. 2

Schematic diagrams of the spatio-temporal meteorological drought characteristics using SPI

SPI estimated based on the cumulative probability function as follows (Eq. 4):
$$ G(x) = \int\limits_{0}^{x} {g(x){\text{d}}x = \tfrac{1}{{\beta^{\alpha } \varGamma \left( \alpha \right)}}\int\limits_{0}^{x} {x^{\alpha - 1} e^{ - x/\beta } } } {\text{d}}x, $$
(4)
where G(x) = cumulative probability of the observed rainfall
US-NDMC (2016) indicated that if the actual rainfall becomes low probability of the cumulative probability function, then the area is under the prevalence of drought; otherwise, the situation reflects wet event. Furthermore, the gamma function is undefined for x = 0 and rainfall may have zero value (Gidey et al. 2018). In this study, the cumulative probability function was analyzed as follows (Eq. 5):
$$ H(x) = q + (1 - q)G(x), $$
(5)
where q = probability of zero.
The fitted rainfall record was transformed into a normal distribution, so that mean SPI is zero and the standard deviation is one. To analyze SPI, the long-term mean seasonal rainfall data of the study area was computed by summing the whole record divided by the number of measurements in the last three decades (Eq. 6):
$$ \bar{x} = \frac{1}{n} \times \sum\limits_{i = 1}^{n} {x_{i} } , $$
(6)
where \( \bar{x} \) = mean value, n = number of observations in the data set or years, and xi = the value of ith timescale being averaged.
The standard deviation was also computed to measure the spread of rainfall as follows (Eq. 7):
$$ \sigma = \sqrt {\frac{1}{n}\sum\limits_{i = 1}^{n} {(x_{i} - \bar{x})^{2} } } , $$
(7)
where σ = standard deviation, xi = each value of data sets, \( \bar{x} \) = mean of the data sets, n = total number of observations, and ∑ = sum of all data sets.
A model was then established to quantify and interpret the values of SPI in the ArcGIS 10.4.1 as follows (Table 2) (Eq. 7):
$$ {\text{SPI}} = {{\left( {p_{i} - \bar{p}} \right)} \mathord{\left/ {\vphantom {{\left( {p_{i} - \bar{p}} \right)} \sigma }} \right. \kern-0pt} \sigma }, $$
(8)
where pi = current rainfall, \( \bar{p} = \) long-term mean, and σ = the standard deviation of the long-term record.
Table 2

Classification of meteorological drought by SPI

(source: McKee et al. 1993; Paulo and Pereira 2006; Rossi et al. 2007)

SPI

Drought classification

Severity

Symbol

≥ 2

Extremely wet

EW

1.5 to 1.99

Very wet

VW

1.0 to 1.49

Moderately wet

MW

0.99 to 0

Mildly wet

MiW

0 to − 0.99

Mild drought

MiD

− 1.0 to − 1.49

Moderate drought

MD

− 1.5 to − 1.99

Severe drought

SD

≤ − 2.0

Extreme drought

ED

2.3.4 Analysis of Drought Duration (L), Magnitude (M), Intensity (I), and Frequency

Gidey et al. (2018) indicated that drought duration could be expressed by evaluating the number of repeated drought events. In this study, the drought duration was investigated based on the number of successive drought incidence days or periods. Drought magnitude is also explained by the summation of rainfall deficits over a particular period. The magnitude of drought was assessed based on McKee et al. (1993) as follows (Eq. 9):
$$ M = - \left( {\sum\limits_{j = 1}^{x} {SPI_{ij} } } \right), $$
(9)
where M =  drought magnitude, x = number of drought months, j =  drought months, and SPIij = SPI value in ith timescale and jth year.
The intensity of drought in the study area from the period 1983–2015 was also statistically analyzed as follows (Eq. 10):
$$ I_{j} = {{M_{j} } \mathord{\left/ {\vphantom {{M_{j} } {L_{j} }}} \right. \kern-0pt} {L_{j} }}, $$
(10)
where Ij = intensity, Mj = drought magnitude, and Lj = drought duration.
Furthermore, the frequency of drought incidence was assessed to measure the regularity of drought events or return period as follows (Eq. 11):
$$ f = \left( {{N \mathord{\left/ {\vphantom {N T}} \right. \kern-0pt} T}} \right) \times 100, $$
(11)
where f = percentage of drought relative frequency, N = number of drought epochs, and T =  total number of observed years (in this case it is 33 years).

3 Results and Discussion

3.1 Seasonal Rainfall Variability

Analysis of the seasonal rainfall variability in the arid and semi-arid regions (e.g., the study area) provides good insight to deal with the seasonal droughts in detail (Gidey et al. 2018). Figure 3 presents the seasonal rainfall condition at each district for the period 1983–2015. The results show that all districts received low and irregular rainfall, leading to disastrous droughts. This study observed that during the last 33 years, the maximum and minimum recorded rainfall was 605, 46 mm in the districts of Hintalo Wejirat and Ofla, respectively. The years 1984, 1987, and 2015 were observed as the worst and/or driest period in the area. Conway (2000) noticed that this driest year recorded was due to low rainfall events. Similarly, this study found that during the years 1983, 1984, 1985, 1987, 1989, 1990, 1991, 1993, 1997, 2002, 2004, 2008, 2009, 2014, and 2015, the areas received a maximum of 382 and minimum of less than 212 mm, respectively (Fig. 3).
Fig. 3

Mean seasonal rainfall in the study area from 1983 to 2015

Furthermore, this study observed that there is very high rainfall variability that ranged between 28.14 and 42.32% in the study area. Table 3 clearly shows that the variability of rainfall in the area was different from district to district. For example, in the district of Ofla, a higher rainfall variation which was estimated at 42.32% was observed. However, lower rate of variability (28.14%) than the other parts of the study area was also observed in the district of Hintalo Wejirat (Table 3). World Bank (2006) reported that rainfall variability may drastically affect the recharge of groundwater in some areas resulting in failure of wells. As a result, the incidences of meteorological drought impact could be severe and affect the production and productivity level of the rain-fed agriculture in the area. The ultimate influencing factors of rainfall variability in Northern Ethiopia in general and the study area in particular were due to El Niño–Southern Oscillation events, which are associated with lower than average rainfall (Conway 2000). Alemayehu and Bewket (2016) showed that the frequent droughts in the area during the last 30 years were due to the limited rainfall arising from the El Niño event. Abebe (2014) also observed the occurrences of drought in the northern Ethiopian highlands caused by high rainfall variability in both spatial distributions and timing. Livada and Assimakopoulos (2007) also found that the root cause of drought that occurred in Greece was due to the lack of adequate rainfall. This study also found that the drought incidence was attributable to the irregular behavior of monsoon rainfall, unreliable distribution and the effect of the Danakil depression, which is one of the driest, hottest and lowest places on the earth.
Table 3

Seasonal rainfall variability in the study area over the period 1983–2015

District

Rainfall (mm)

Min

Max

Mean

SD

CV

CV  %

Yalo

106.00

546.87

320.90

109.79

0.34

34.21

Megale

11.53

604.22

372.41

128.51

0.35

34.51

Gulina

103.83

446.56

273.61

90.21

0.33

32.97

Raya Azebo

90.19

516.08

316.29

114.29

0.36

36.13

Alamata

93.27

481.60

294.02

104.00

0.35

35.37

Hintalo Wejirat

162.37

605.25

401.16

112.89

0.28

28.14

Kobo

133.67

517.89

306.08

100.80

0.33

32.93

Endamehoni

59.62

486.99

274.67

111.18

0.40

40.18

Ofla

45.67

424.88

231.64

98.03

0.42

42.32

Alaje

73.37

517.95

312.62

113.59

0.36

36.34

Gidan

84.29

508.84

278.57

107.80

0.39

38.70

3.2 Meteorological Drought Conditions in the Study Area

The Standardized Precipitation Index (SPI-3) model indicated that the spatio-temporal meteorological drought incidence varied from year to year (Figs. 4, 5a–k). This probably occurred when the SPI-3 result was negative and ended when it is positive. Hence, this study found that the years 1986, 1992, 1995, 1996, 2000, 2005, 2011, and 2012 were moderately wet epochs in all districts in the last 33 years. Similarly, the years 1988, 1994, 1998, 1999, 2001, 2003, 2006, 2007, and 2010 were wet periods (Fig. 5a–k). The year 1998 was wet across all districts. As a result, the areas were free of drought incidence. However, drought intermittently struck the districts during the 1983, 1984, 1985, 1987, 1989, 1990, 1991, 1993, 1997, 2002, 2004, 2008, 2009, 2014, and 2015 period (Fig. 5a–k).
Fig. 4

A spatio-temporal meteorological drought condition in the study area over the period 1983–2015. Note*: Afa_1: Megale, Afa_2: Yalo, Afa_3: Gulina, Tig_1: Alaje, Tig_2: Alamata, Tig_3: Hintalo Wejirat, Tig_4: Ofla, Tig_5: Endamehoni, Tig_6: Raya Azebo, Amh_1: Gidan, Amh_2: Kobo

Fig. 5

a–k

Meteorological drought condition in the study area based on SPI–3

3.3 Historical Meteorological Drought Onset (O), Offset (E), Duration (L), Magnitude (M), and Intensity (I)

Tables 4, 5, and 6 show that the onset, offset, duration, magnitude, and intensity of meteorological drought in the study area during the period 1983–2015. The results indicated that drought struck extensively for a maximum of 3 years between the periods 1983–1985, 1989–1991 and 2013–2015, respectively (Tables 4, 5, 6). For example, in the lowland area (Table 4), the onset, offset, duration of meteorological drought were 1–3 years throughout the study period. Similarly, the mid- and highlands also experienced comparable shocks (Tables 5, 6). There is no difference in the duration of drought periods. During this study period, significant impacts on the livestock, livestock products, and humans were observed throughout all agro-ecological zones. Particularly, the 1983–1985 droughts claimed the lives of thousands of smallholder farmers because the impact continued for consecutive 3 years. Characterizing the major properties of meteorological drought may give in-depth information for implementing better monitoring, adaptation and mitigation strategies.
Table 4

Historical meteorological drought onset (O), offset (E), duration (L), magnitude (M), and intensity (I) in lowlands of the study area

O

E

Yalo

Gulina

Megale

L

M

I

L

M

I

L

M

I

1983

1985

3

− 3.50

− 1.17

3

− 3.39

− 1.13

3

− 3.59

− 1.20

1987

1987

1

− 1.06

− 1.06

1

− 0.86

− 0.86

1

− 1.35

− 1.35

1989

1991

3

− 2.60

− 0.87

3

− 3.08

− 1.03

3

− 2.19

− 0.73

1993

1993

1

− 0.95

− 0.95

1

− 0.96

− 0.96

1

− 1.03

− 1.03

1997

1997

1

− 0.87

− 0.87

1

− 0.95

− 0.95

1

− 0.91

− 0.91

2002

2002

1

− 0.31

− 0.31

1

− 0.39

− 0.39

1

− 0.26

− 0.26

2004

2004

1

− 0.54

− 0.54

1

− 0.51

− 0.51

1

− 0.48

− 0.48

2008

2009

2

− 0.66

− 0.33

2

− 0.47

− 0.24

2

− 0.80

− 0.40

2013

2015

3

− 1.59

− 0.53

3

− 1.57

− 0.52

1

− 0.09

− 0.09

  

2

− 1.56

− 0.78

Table 5

Historical meteorological drought onset (O), offset (E), duration (L), magnitude (M), and intensity (I) in the midlands of the study area

O

E

Raya Azebo

Hintalo-Wejirat

Kobo

Alamata

L

M

I

L

M

I

L

M

I

L

M

I

1983

1985

3

− 0.37

− 0.12

3

− 3.39

− 1.13

3

− 3.23

− 1.08

3

− 3.58

− 1.19

1987

1987

1

− 1.17

− 1.17

1

− 1.63

− 1.63

1

− 0.78

− 0.78

1

− 1.04

− 1.04

1989

1991

3

− 2.55

− 0.85

3

− 1.55

− 0.52

3

− 2.95

− 0.98

3

− 2.78

− 0.93

1993

1993

1

− 0.94

− 0.94

1

− 0.90

− 0.90

1

− 0.85

− 0.85

1

− 0.90

− 0.90

1997

1997

1

− 0.86

− 0.86

1

− 0.62

− 0.62

1

− 0.85

− 0.85

1

− 0.81

− 0.81

2002

2002

1

− 0.30

− 0.30

1

− 0.28

− 0.28

1

− 0.26

− 0.26

1

− 0.20

− 0.20

2004

2004

1

− 0.58

− 0.58

1

− 0.58

− 0.58

1

− 0.61

− 0.61

1

− 0.60

− 0.60

2008

2009

2

− 0.72

− 0.36

2

− 0.65

− 0.33

1

− 0.47

− 0.47

2

− 0.71

− 0.36

2014

2015

2

− 1.53

− 0.77

1

− 0.12

− 0.12

1

− 0.19

− 0.19

3

− 1.62

− 0.54

− 

− 

2

− 1.68

− 0.84

2

− 1.29

− 0.65

   
Table 6

Historical meteorological drought onset (O), offset (E), duration (L), magnitude (M), and intensity (I) in the highlands of the study area

O

E

Endamehoni

Ofla

Alaje

Gidan

L

M

I

L

M

I

L

M

I

L

M

I

1983

1985

3

− 3.79

− 1.26

3

− 3.80

− 1.27

3

− 3.90

− 1.30

3

− 3.53

− 1.18

1987

1987

1

− 0.96

− 0.96

1

− 0.98

− 0.98

1

− 1.25

− 1.25

1

− 0.85

− 0.85

1989

1991

3

− 3.04

− 1.01

3

− 3.18

− 1.06

3

− 2.60

− 0.87

3

− 3.00

− 1.00

1993

1993

1

− 0.96

− 0.96

1

− 0.86

− 0.86

1

− 0.97

− 0.97

1

− 0.91

− 0.91

1997

1997

1

− 0.86

− 0.86

1

− 0.78

− 0.78

1

− 0.71

− 0.71

1

− 0.86

− 0.86

2002

2002

1

− 0.19

− 0.19

1

− 0.14

− 0.14

1

− 0.23

− 0.23

1

− 0.18

− 0.18

2004

2004

1

− 0.56

− 0.56

1

− 0.56

− 0.56

1

− 0.48

− 0.48

1

− 0.66

− 0.66

2008

2009

2

− 0.71

− 0.36

2

− 0.69

− 0.35

2

− 0.48

− 0.24

2

− 0.80

− 0.40

2013

2015

3

− 1.52

− 0.51

3

− 1.65

− 0.55

1

− 0.14

− 0.14

4

− 1.79

− 0.45

        

2

− 1.46

− 0.73

   

3.4 Meteorological Drought Frequency, Severity, and Spatial Extent in the Study Area

Based on Fig. 5a–k, the severity, frequency, and spatial extent of the meteorological drought event are presented in Table 7. The result shows that meteorological drought is a regular incidence in all districts of the study area. The frequency in dry spells, as well as drought incidence, exacerbates crop failure and food insecurity and poverty (Awulachew et al. 2007). However, in the last 33 years, the study areas were not hit by extreme drought (ED) and extreme wet (EW) conditions. The SPI-3 result indicates that the year 1984 was identified as a severe drought period than the 2015 drought event that covering nearly 14,532 km2 spatially (Table 7). During this period, millions of people and livestock were faced with acute food shortages. This drought event was bad because it claimed the lives of thousands of humans and animals. Thousands of children were also left malnourished, and migrated in large numbers to neighboring countries such as Sudan (Assefaw 1993). This study also shows that the primary cause of drought in this period was the shortfall of the main rainy season. Besides, the poor drought mitigation policies and measures undertaken by the ruling military government (the so-called ‘Derg’) aggravated the impact of drought on people and livestock. Further, the study found that one severe drought occurred in all districts of the study area except in Hintalo Wejirat. Four moderate droughts also hit all districts except Gulina, Hintalo Wejirat, and Kobo (Table 7). In the entire areas, about 10–12 mild droughts were observed. Hence, the frequency of drought events in the study area was found to occur once in every 2–3 years.
Table 7

Frequency, severity, and spatial extent of meteorological drought incidence from 1983 to 2015

District

Spatial extent (km2)

Drought frequency and severity

Extreme drought (ED)

Severe drought (SD)

Moderate drought (MD)

Mild drought (MiD)

≤− 2

%

− 1.5 to − 1.99

%

− 1.0 to − 1.49

%

0 to − 0.99

%

Yalo

1716.5

0

0

1

3

4

12

10

30

Megale

1543.5

0

0

1

3

4

12

11

33

Gulina

1149.7

0

0

1

3

3

12

11

33

Raya Azebo

1782.7

0

0

1

3

4

12

10

30

Alamata

725.5

0

0

1

3

4

12

11

33

Hintalo-Wejirat

1933.1

0

0

2

7

3

9

12

36.4

Kobo

1942.7

0

0

1

3

1

3

12

36.4

Endamehoni

634.8

0

0

1

3

4

12

11

33

Ofla

1101.3

0

0

1

3

4

12

11

33

Alaje

773.2

0

0

1

3

4

12

11

33

Gidan

1229.2

0

0

1

3

4

12

12

36.4

4 Conclusions

In this research, the spatio-temporal meteorological drought onset, offset, duration, magnitude, intensity, frequency, severity, and spatial extent are examined using the Standardized Precipitation Index (SPI) at 3-month timescale. The long-term analysis of rainfall indicated that all sites received minimal and irregular rainfall, leading to climate extremes such as droughts. The study area experienced very high rainfall variability ranging between 28.14 and 42.32% in the study area. The SPI-3 month timescale model reveals that the study area was stricken by mild-to-severe drought once in every 2–3 years. This drought regularity is observed at a higher rate than the previous decades. This study, therefore, realized that meteorological drought is a regular incidence in all districts of the study area. However, neither extreme drought (ED) nor extreme wet (EW) conditions were found. The year 1984 identified as a severe drought period in all parts of the study area covered nearly 14532 km2 spatial extent. Similarly, the study found that four moderate droughts hit all districts except Gulina, Kobo, and Hintalo Wejirat, while in the entire areas about 10–12 mild droughts incidence was observed. This study found that SPI is the most crucial drought index for both local and global meteorological drought-related studies. Drought is severe in the study area and it is anticipated to continue in the future due to shortages of adequate rainfall. This study could provide better insight for policymakers on regional drought monitoring and early warning system to improve the existing monitoring, mitigation and adaptation strategies and the ongoing water harvesting activities in order to reduce the impact of climate extremes and drought. Further, it will support to enhance the livelihood of smallholder farmers through implementing drought-tolerant crop varieties in quality and quantity, substantial irrigation practices, and climate-based crop insurance schemes.

Notes

Acknowledgements

We are grateful for the research fund provided by Mekelle University and the OSF–ACCAI project of Mekelle University, Ethiopia. The main author is thankful for the PhD scholarship given by the TreccAfrica II project. The authors would also like to acknowledge the National Meteorological Agency of Ethiopia (NMA) for providing the climatic data. We would also like to thank the constructive feedback of the two anonymous reviewers and the editor.

Compliance with ethical standards

Conflict of interests

On behalf of all authors, the corresponding author states that there is no conflict of interest.

References

  1. Abebe BA (2014) Modeling the Effect of Climate and Land Use Change on the Water Resource in Northern Ethiopia (Doctoral dissertation, Freie Universität Berlin)Google Scholar
  2. AghaKouchak A, Farahmand A, Melton FS, Teixeira J, Anderson MC, Wardlow BD, Hain CR (2015) Remote sensing of drought: progress, challenges and opportunities. Rev Geophys 53(2):452–480CrossRefGoogle Scholar
  3. Alemayehu A, Bewket W (2016) Local climate variability and crop production in the central highlands of Ethiopia. Environ Dev 19:36–48CrossRefGoogle Scholar
  4. Assefaw T (1993) Participatory relief management: the experience of the relief society of tigray. Refug Can J Refug 12(8):23–25Google Scholar
  5. Awulachew SB, Yilma AD, Loulseged M, Loiskandl W, Ayana M, Alamirew T (2007) Water resources and irrigation development in Ethiopia, vol 123. Iwmi, ColomboGoogle Scholar
  6. Ayenew T, GebreEgziabher M, Kebede S, Mamo S (2013) Integrated assessment of hydrogeology and water quality for groundwater-based irrigation development in the Raya Valley, northern Ethiopia. Water Int 38(4):480–492CrossRefGoogle Scholar
  7. Belal AA, El-Ramady HR, Mohamed ES, Saleh AM (2014) Drought risk assessment using remote sensing and GIS techniques. Arab J Geosci 7(1):35–53CrossRefGoogle Scholar
  8. Bonaccorso B, Bordi I, Cancelliere A, Rossi G, Sutera A (2003) Spatial variability of drought: an analysis of the SPI in Sicily. Water Resour Manage 17(4):273–296CrossRefGoogle Scholar
  9. Conway D (2000) Some aspects of climate variability in the north east Ethiopian Highlands-Wollo and Tigray. Sinet Ethiop J Sci 23(2):139–161CrossRefGoogle Scholar
  10. Gidey E, Dikinya O, Sebego R, Segosebe E, Zenebe A (2017) Modeling the spatio–temporal dynamics and evolution of land use and land cover (1984–2015) using remote sensing and GIS in Raya, Northern Ethiopia. Model Earth Syst Environ 3(4):1285–1301CrossRefGoogle Scholar
  11. Gidey E, Dikinya O, Sebego R, Segosebe E, Zenebe A (2018) Predictions of future meteorological drought hazard (~ 2070) under the representative concentration path (RCP) 4.5 climate change scenarios in Raya, Northern Ethiopia. Model Earth Syst Environ.  https://doi.org/10.1007/s40808-018-0453-x CrossRefGoogle Scholar
  12. Kogan F, Guo W (2016) Early twenty-first-century droughts during the warmest climate. Geomat Nat Hazards Risk 7(1):127–137CrossRefGoogle Scholar
  13. Kogan F, Adamenko T, Guo W (2013) Global and regional drought dynamics in the climate warming era. Remote Sens Lett 4(4):364–372CrossRefGoogle Scholar
  14. Livada I, Assimakopoulos VD (2007) Spatial and temporal analysis of drought in Greece using the standardized precipitation index (SPI). Theor Appl Climatol 89(3–4):143–153CrossRefGoogle Scholar
  15. McKee TB, Doesken NJ, and Kleist J (1993) The relationship of drought frequency and duration to timescales. In Proceedings of the 8th Conference on Applied Climatology. Boston, MA, USA: American Meteorological Society, 17(22):179–183Google Scholar
  16. Mishra AK, Singh VP (2010) A review of drought concepts. J Hydrol 39(1):202–216CrossRefGoogle Scholar
  17. Paulo AA, Pereira LS (2006) Drought concepts and characterization: comparing drought indices applied at local and regional scales. Water Int 31(1):37–49CrossRefGoogle Scholar
  18. Peters AJ, Walter-Shea EA, Ji L, Vina A, Hayes M, Svoboda MD (2002) Drought monitoring with NDVI–based standardized vegetation index. Photogramm Eng remote Sens 68(1):71–75Google Scholar
  19. Rossi G, Bonaccorso B, Vega T (eds) (2007) Methods and tools for drought analysis and management, vol 62. Springer Science and Business Media, Berlin. ISBN 978-1-4020-5923-0Google Scholar
  20. Şen Z, Al-Harithy S, As-Sefry S, Almazroui M (2017) Aridity and risk calculations in saudi arabian wadis: Wadi Fatimah case. Earth Syst Environ 1(2):26CrossRefGoogle Scholar
  21. Shamsipour AA, Zawar-Reza P, Alavi Panah SK, Azizi G (2011) Analysis of drought events for the semi–arid central plains of Iran with satellite and meteorological based indicators. Int J Remote Sens 32(24):9559–9569CrossRefGoogle Scholar
  22. Svoboda M, Hayes M, Wood D (2012) Standardized precipitation index user guide. World Meteorological Organization, GenevaGoogle Scholar
  23. Tabari H, Taye MT, Willems P (2015) Statistical assessment of precipitation trends in the upper Blue Nile River basin. Stoch Env Res Risk Assess 29(7):1751–1761CrossRefGoogle Scholar
  24. US National Drought Mitigation Centre (US–NDMC) 2016. The Standardize Precipitation Index. http://www.drought.unl.edu/monitor/spi/program/SPI_Zimbabwe.doc, Accessed on 06/06/2016
  25. Wang PX, Li XW, Gong JY, Song C (2001) Vegetation temperature condition index and its application for drought monitoring. In: Geoscience and Remote Sensing Symposium, 2001. IGARSS’01. IEEE 2001 International. IEEE, 1:141–143Google Scholar
  26. Wilhite DA (1996) A methodology for drought preparedness. Nat Hazards 13(3):229–252CrossRefGoogle Scholar
  27. World Bank (2006). Ethiopia: managing water resources to maximize sustainable growth. Report no. 36000-ETGoogle Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  • Eskinder Gidey
    • 1
    • 2
    • 3
  • Oagile Dikinya
    • 1
  • Reuben Sebego
    • 1
  • Eagilwe Segosebe
    • 1
  • Amanuel Zenebe
    • 2
    • 3
  1. 1.Department of Environmental ScienceUniversity of BotswanaGaboroneBotswana
  2. 2.Land Resource Management and Environmental Protection, Mekelle UniversityMekelleEthiopia
  3. 3.Institute of Climate and Society, Mekelle UniversityMekelleEthiopia

Personalised recommendations