Climate and Agriculture : Empirical evidence for Countries and Agroecological Zones of the Sahel 

Get Complete Project Material File(s) Now! »

Climate change: definitions, measures and models

In this first section, we discuss and define the climate and global warming, their measures and the models used to measure climate and climate change.

Definitions

Climate

In order to link climate change to the economy, we first need to understand climate and the phenomenon of climate change Stern (2008). In the scientific literature and public discourses, there is no single definition of climate and global warming. Climate experts are not unanimous on the definition of climate, global warming, climate trends or other fluctuation (Werndl, 2016). However, the issue at stake is important as a poor definition of these terms leads to confusion and incomprehension of the climate system. We therefore provide here some key aspects that can provide a conceptual definition of climate and climate change.
Pachauri et al. (2014) define the climate, climate change and the various phenomena that can result. The intergovernmental body provides a definition in the strict sense of the climate, according to which « climate generally refers to average time or more precisely to a statistical description based on averages and variability of relevant quantities over periods ranging from a few months to thousands, the standard period, as defined by the World Meteorological Organization, is 30 years. These quantities are usually surface variables such as temperature, precipitation height and wind. In a broader sense, climate refers to the state of the climate system, including its statistical description ». According to this definition, we argue that the global climate is a « statistical distribution » of all terrestrial atmospheric conditions in the world over a period of time.
The traditional definition of climate is that of the statistical properties of the meteorological conditions observed at the time and at the place of the year, and these statistical properties are determined from observations made over a certain reference period (Stone et al., 2009). Indeed, the climate varies from one place to another, depending on latitude, distance to the sea, vegetation, presence or absence of mountains or other geographical factors. The climate also varies over time; from one season to another, from one year to another, from one decade to another or from a much longer time scale, such as Ice Age (Lead, 2000).
To these definitions of climate, we must also add that of the weather which is defined by Lead (2000) as « the fluctuating state of the atmosphere that surrounds us, characterized by temperature, wind, precipitation, clouds and other meteorological elements ». Thus, the knowledge of the weather makes it possible to have an idea on the distribution of the climate in which we live.
The Köppen climate classification system lists five types of climate based on monthly and annual averages of precipitation and temperature. Each type of climate is indicated by a capital letter: wet humid climates (A), dry climates (B), mid-latitude climate with mild winters (C), wet climates in mid-Latitude with cold winters (D) and polar climates (E) with very cold winters and summers (Pidwirny and Jones, 2006).
The spatial and temporal aspects must be taken into account Stone et al. (2009) in the definition of climate. When discussing the climate of a country or region, we need to consider climate variables or factors on the same scale. In general, they are dynamic meteorological variables such as surface air temperature or surface pressure describing the state of the atmosphere at a given time (Werndl, 2016). In other words, climate is a set of weather conditions, a distribution of climatic variables that can appear for a certain configuration of the climate system (Bradley et al., 2017).
A distinction must be made between weather and climate. Allen (2003) quotes Edward Lorenz (indefinite) who states that « climate is what you expect, time is what you get ». Allen (2003) argues that climate is « time » and that statistics are defined by the statistician as the « expected time » and its variability for a given time, taking into account all the properties of the oceanic atmospheric system, emissions Current greenhouse gas emissions, solar activity, etc.
Based on these definitions, variables of interest are dynamic climatic variables (Stone et al., 2009). On the other hand, in the literature, several variables can explain climate, for example the temperature of the ocean (Werndl, 2016). The list of climatic variables is long, it can group the variables describing flora and fauna, but they are not often taken into account in the literature . Furthermore, temperature and precipitation are also relevant variables that may explain the climatic situation of a given geographic or geographic area.
Figure 1.1 displays Köppen’s climate classification. The climatic conditions can be different at the level of a continent, a country and even at the level of a country. They vary temporarily and spatially. Thus, the climate system is the sum of all climates in each geographical area. For example, the tropical climate is characterized by high and constrasted temperatures, which vary according to the season, the temperature in summer is around 23 C and the winter temperature is around 35 C (Paturel et al., 1997). Climatic conditions may be unfavorable for desert areas (African and Arab countries) with temperatures up to 46 C and rainfall often low and abnormal (Al-Mebayedh, 2013). Being a long-term phenomenon, climate can vary according to regions, countries and continents. It is important to see how it can change over a period so we define climate change in the next section.
Being a global phenomenon, global warming is generally characterized by an increase in the average temperature of the oceans and the atmosphere. In other words, this would result from a sharp increase in the concentration of greenhouse gases (NAZA, 1998), such as carbon dioxide, methane and nitrogen dioxide. Indeed, these gases are responsible for climate change (Muller, 2013).
Climate change is also considered as the variation in the state of the climate, which is detectable by statistical tests (Pachauri et al., 2014; Werndl, 2016). This variation translates into changes in the mean and / or variability of climatic properties over a long period of time, decades or more. In other words, climate change can be assimilated to the variation over time of meteorological factors (Stone et al., 2009).
In its first article, the UNFCCC defines climate change as « climate change that is directly or indirectly attributed to human activity that changes the composition of the global atmosphere and adds to the natural climatic variability observed under comparable conditions ». According to this definition, the UNFCCC distinguishes between climate change caused by human activities on the atmosphere and those caused by natural problems.
Indeed, the natural greenhouse effect creates habitable climatic conditions in which humankind can aspire to live in relatively benign conditions, otherwise the earth would be a very icy and unbearable place. An increased greenhouse effect, however, refers to the possible rise in the average temperature of the earth’s surface that is greater than that due to the natural greenhouse effect due to an increase in greenhouse gas concentrations caused by human activities (NAZA, 1998).
Such global warming as a result of an increase in the greenhouse effect would likely lead to other, sometimes harmful changes in the climate for example, changes in precipitation, thunderstorms and the level of the oceans (NAZA, 1998).
For Stone et al. (2009), climate change refers to any change in climate whether it is forced naturally or anthropologically.
Werndl (2016) lists five criteria for a rigorous definition of climate change. The first criterion suggests that the definition of climate must be empirically applicable, i.e. the definition must allow an estimation of the past and future climate. The second criterion indicates that the definition of climate must correctly classify the different climate over different periods. The third criterion stresses that the climate must not depend on our knowledge, i.e. it does not take into account the speculations made on the climate. The fourth criterion indicates that the definition of climate must apply to the past, present and future. Finally, the fifth criterion emphasizes that a definition of climate must be mathematically well defined. In general, the criteria defined must lead to a definition that must be empirically applicable and that takes account of past and future climate. This definition should also make it possible to classify the different types of climate over time. Thus, Werndl (2016) defines climate as a finite distribution over time resulting from the regime of variable external conditions. In other words, actual external conditions over a period of time are subject to a certain regime of varying external conditions. This definition is new and is not yet shared in the climate literature. With this in mind, Werndl (2016) speaks of climate change when there are different climates for two successive periods of time, and that there may be external and internal climate change due to different initial values. For Werndl (2016), this definition is specific because it is empirically applicable, the actual climate system is subject to a certain regime of variable external conditions over a sufficiently long period, so the climate of this period coincides with the distribution in time of the actual evolution of climate variables. This definition is also unique because it makes it possible to make an immediate link with the observations.
Focusing on Africa and the Sahel, climate change is perceived differently. For Ouédraogo et al. (2010), climate change leads to environmental degradation. Thus, Sahelian peasants perceive changes in precipitation through its direct effects on soils and vegetation cover. For instance, according to Ouédraogo et al. (2010), climate change in Burkina Faso is perceived by farmers through rainfall changes as agricultural activities depend on precipitation. Climate change results in decreased and increasing irregularities in rainfall, a deregulation of the winter season, and a high frequency of drought (Ouédraogo et al., 2010).
Summarizing, climate change measures depend on the climate variables used, the perception of farmers and people with activities dependent on climate change. If climate change can be observed over a long period of time, the local population aged in a given region is better placed to testify. The climate is variable over time, it can be explained by several indicators that can translate this variation on the different systems that are linked to the climate to a certain extent.

Measures

Climate change is a disruption of the climate system and climate measurement involves comparing the results of the climate variable to its equilibrium value. It is measured by the standard deviation or the average absolute deviation of the distribution of a variable from its mean or long-term trend. Moreover, it should be noted that the standard deviation weighs and evaluates better the extreme events (Badolo and Kinda, 2014). Climate change indicators may vary according to geographic and climatic zones. At the global level, IPCC observations are made through land temperature, ocean heat content, sea level, and atmospheric water value. Recent observations have shown that these indicators are high, and are considered by scientists to be evidence of climate change (Smith et al., 2014). The high frequency of droughts over the last 50 years can explain and also measure the effects of climate change on the economies of the five continents. Based on the Center for Research on the Epidemioligy of Disasters database, Masih et al. (2014) report that the world experienced 642 droughts worldwide during the period from 1900 to 2013. These events killed over 11 million people, affected more than two billion and estimated economic damage to more than 100 million for the whole world.
Klos et al. (2015) conducted surveys of natural resource professionals to list biophysical indicators to obtain indicators or local factors to assess climate change in Idaho and provide relevant information to decision makers to measure climate change. They argue that changes in water resources and risk of forest fires are the most important issues for the professionals interviewed. They listed indicators that included direct climate measurements (air temperature, rainfall, stream temperature, snowpack, streamflow, drought, plant phenology) and indicators partially influenced by climate (fire disturbance, human activities, species viability and productivity, biotic disturbance, animal phenology). These indicators are classified into two dimensions according to their relationship with the climate and their involvement in the triggering of the climate change process. Other variables that are partially influenced by climate can be strongly controlled by other mechanisms such as land management, ecological stressors (Klos et al., 2015).
Climate change measures may also vary from one area to another depending on climatic characteristics. In the Sahel, agricultural practices are mostly rainfed and climatic variables such as precipitation and temperature can determine climatic variability. Farmers perceive climate change as rainfall decreases because the quantity and distribution of annual and decadal rainfall is highly variable (Branca et al., 2013) and are the main constraints to agricultural development (Sissoko et al., 2011). For example, Eric and Kinda (2016) argue that climate change translates into rainfall variability, drought, floods and extreme temperatures. According to these authors, climate change is measured by the standard deviation of the rate of growth of the water balance (the difference between precipitation and evaporation) and extreme events (drought, floods and extreme temperatures). It should also be noted that, in the Sahel, environmental variables such as desertification and dwindling water supplies are pointed out as threats and it should be added that lack of resources and soil depletion are due to climate change (Heinrigs, 2010). Consequently, changes in rainfall patterns, temperature and / or frequency or severity of extreme events will have direct impacts on crop yields (Sissoko et al., 2011).

Models

Definitions of climate and climate change cover statistical aspects, suggesting the existence of climate models to measure these climatic variations (Stone et al., 2009). Thus, scientists use mathematical and statistical models to analyze the distribution of climate variables. The call to mathematics solves the complex equations derived from these laws. For ?, the characteristics and the particularity of climate processes leave no room for laboratory experiments, the only way is mathematical modeling, hence the complexity of the definition of climate change. This approach is also shared by Hulme et al. (2009) who argue that the climate can not be measured by our senses or by our instruments such as wind and water. They also support that the climate has several interpretations because it contains a physical definition (the Amazonian climate is more humid than that of the Sahara) and cultural (the meaning of the climate of the Sahara is totally different for a bedouin than a berliner). Before enumerating the climatic models developed in the climate literature, we define and describe a climate model. The aim of the section is not to detail climate models in general but we provide a theoretical analysis while evoking the debates built around the modeling of the climate. The IPCC defines a climate model as a numerical representation of the climate system based on the physical, chemical and biological properties of its components, interactions and feedback processes, and the recognition of some of its known properties Pachauri et al. (2014). Currently, there are about 20 climate models that are continually being developed by national modeling centers such as NASA, the UK Met Office and the Beijing Climate Center Bradley et al. (2017). Old climate models were based on moisture and cloud processes, whereas the new models take into account the role of vegetation, forests, grasslands and crops in measuring and controlling the amount of dioxide Carbon in the atmosphere (NAZA, 1998). According to Parker (2006), these models may be competing or compatible, while emphasizing that pluralism in climate modeling combines an ontic competitive pluralism with an integrative pragmatic pluralism. In other words, the ontic dimension concerns the compatibility of hypotheses on what is « the world », whereas the pragmatic dimension concerns the compatibility of models in practice.
Climate models are used as a research tool to study and simulate climate for operational purposes, including monthly, seasonal and interannual climate forecasts (Pachauri et al., 2014). In addition, models can also describe the state of the climate system. For example, the empirical study of climate requires observations and models that provide such data. While the data show that the hottest temperatures have been observed since the 20th century, the models also allow scientists to prove the existence of climate change and to detect the causes of this temperature rise Bradley et al. (2017). In other words, models are relevant tools for characterizing and evoking the causes of climate change.
The models also seek to identify the origin and factors responsible for climate change. Human activities have long been indexed and are considered the main sources of climate change. For that reason, questions about the existence of climate change and its attribution raise problems Bradley et al. (2017), known as the detection problem (Pachauri et al., 2014). Thus, the detection of climate change is defined by the IPCC as a process by which scientists demonstrate that the climate or a climate system has changed, without giving the reason for this change. Moreover, the identified change is detected in the observations if it is established that its probability of occurrence by chance arising only from internal variability is low (Pachauri et al., 2014).
According to Bradley et al. (2017), this definition is inadequate with that of Werndl (2016) which defined climate as a finite distribution over a relatively short period of time. However, this definition is reported by climatologists and scientists to set up statistical tests and hypotheses to detect climatic variation. Climate change is detected when observed values fall outside a predefined range of internal variability (Pachauri et al., 2014). Thus, the tests carried out allow scientists to know whether the detected climate change is due to human activities or other causes. Climate change attribution is a mechanism to assess the contribution of different factors to climate change detected with a statistical confidence level (Pachauri et al., 2014). However, the representation of the climate system can be made differently and by models of different complexity, i.e. they can use a single component or a combination of several components (Pachauri et al., 2014).
For Goosse (2015), climate models should use at least the physical behavior of the atmosphere, ocean and sea ice. In addition, terrestrial carbon cycles, vegetation and Ice cap are also taken into account, thus giving models of the terrestrial system. This analysis is also shared by Parker (2006) who argue that climate scientists use an approach based on a multitude of models to represent the climate system.
According to Pachauri et al. (2014), climate models are differentiated by the number of spatial dimensions, how physical, chemical or biological processes are considered and their empirical parameterizations. For Rial et al. (2004), the Earth’s climate system is non-linear because of the disproportion of inputs and outputs. To this end, climate change corresponds to a sequential process or to episodes where multiple equilibria are the norm. They also consider that the climate system is not really a stationary process, but is subjected to natural and anthropogenic variations in forcing. For Parker (2006), climate models are opposed to one another because of the assumptions made and scientists are unable to announce that one model is better than the other models in a study of future climate change. This difficulty can be explained by the uncertainty that conditions the climate and the difficulties encountered during the assessment of climate models.
To understand and predict climate change, NAZA (1998) proposes the following models. First, socio-economic models that predict the future use of fossil fuels and the use of alternative fuels. These models depend on technology, public policy and social attitudes, economic development, standard of living and the habit of people to resort to energy and chemicals. Then, the psycho-chemical models of the terrestrial system that give an idea of the amount of gas released into the atmosphere. The chemicals and natural processes on the surface of the earth affect the release. Finally, coupled ocean atmosphere models provide information on how the climate system (temperature, humidity, clouds and precipitation) responds to climate change in the composition of the atmosphere (NAZA, 1998; Stone et al., 2009). In addition to these models, others are proposed in the climate literature. We mention here the main ones. Energy balance models (EBMs) or simple models (Stone et al., 2009) were introduced by Budyko (1969) whose objective is to estimate climate change from the energy balance of the earth, considering that the earth as a flat surface with a layer at the top (Bradley et al., 2017). In addition to their simplicity, these models only provide average global values for the calculated variables and provide a good qualitative understanding of the greenhouse effect (Bradley et al., 2017).
Among the models, there are also models of intermediate complexity (EMICs), which are called models of reduced complexity (Stone et al., 2009). They complete the niche between the EBMs and the GCMs. They are simple (Stone et al., 2009) but they always include a geographical representation of the earth (Goosse, 2015; Stone et al., 2009). In other words, these models provide more than averages over the entire land or more vague areas. Furthermore, they include much more degrees of freedom than EMB models (Goosse, 2015) and are used in paleo-climatic applications because of their effectiveness (Stone et al., 2009). On the other hand, EMICs have parameters that are difficult to adjust to reproduce the observed characteristics of the climate system, for example some simpler models, for which reason the level of approximation is chosen considerably between the different EMICs (Stone et al., 2009).
In addition, there are also general circulation models (GCMs) that provide a more accurate and complex description of the climate system (Goosse, 2015). Currently, they are the most commonly used for projections reported in IPCC assessment reports (Stone et al., 2009).
In previous years, the GCMs models only consisted of a representation of the atmosphere, the land surface, sometimes ocean circulation, and a more simplified version of sea ice (Stone et al., 2009). Currently, they take into account several components, and many newly developed models include sea ice, carbon cycle, ice sheet dynamics and atmospheric chemistry (Stone et al., 2009).
All climate models require several components, and consideration of these components depends on the modeler’s goal (Stone et al., 2009). The main components of global studies can be the atmosphere, ocean, sea ice, land surface, marine biogeochemistry, ice sheet and possible coupling between components and Models of terrestrial systems. The effects of climate change are diverse and affect all of humanity. Human, animal and plant species are the main victims of climate problems. They also affect economies and threaten human activities. After elucidating the models assessing climate change, we must identify the main causes and consequences of global warming.
1.3 Climate change and economic activities
In this section, we explain the link between climate change and human activities. First, we explain how human activities and nature affect the climate. Next, we address the effects of this change on human economic activities. Climate change has a significant impact on human health, economic activity and the environment. Finally, we address the channels through which climate change affects a country’s economy by delaying economic activity in countries whose agriculture represents an important economic resource.

READ  The wave equation with a harmonic potential 

Causes of climate change

The causes of climate change can be natural (the presence of greenhouse gases in the atmosphere) and / or anthropic (land-use change, the combustion of fossil fuels, sulfate aerosols and black carbon). The discussion will focus on Africa and the Sahel, because according to our research, Africa is the only continent with a high number of droughts and heavy human consequences.

Natural Causes of Climate Change

Climate change can be caused by natural factors such as the El Niño phenomenon, droughts, floods, volcanic eruptions and other natural factors that could alter the climate system. The natural causes of climate change are often caused by water. Scientists are certain of the link between climate change and rainfall variability (Stern, 2008). For example, the majority of the disastrous effects of climate change are manifested through water: glacier melt and floods, droughts, storms, rising ocean levels. Since the 1950s, many changes have been unprecedented for millennia. The atmosphere and ocean have warmed, snow and ice cover has decreased, and sea levels have risen (Pachauri et al., 2014). However, the causes may vary from one geographical area to another.
Natural causes causing drought in Africa have been studied in several studies (Caminade and Terray, 2010; Dai, 2011, 2013; Giannini et al., 2008; Manatsa et al., 2008; Nicholson et al., 2000). These authors have focused on anthropogenic factors to try to explain the causes of drought in Africa. For them, climate change, aerosol emissions, land-use practices, and land-atmosphere interactions are mechanisms for drought induction (Desboeufs et al., 2010; Hwang et al., 2013; Masih et al., 2014). There are also other factors that can lead to drought conditions such as El Niño-Southern Oscillation (ENSO) and SST due to their strong influence on the continent as a whole. These two factors and land-atmospheric feed-back are the main factors that determine and explain rainfall variability in Africa (Nicholson et al., 2000). The atmosphere is a determining factor in explaining the situation of the terrestrial climate system. For Nicholson et al. (2000), these factors alone or in combination can change atmospheric dynamics and patterns of circulation. For example, they cause changes in Hadley and Walker flows or higher-level stream flows. The results coincide with the work of Rouault and Richard (2005) who have done similar studies but using data on the Normalized Precipitation Index from 1900 to 2004 in South Africa. The results obtained show a strong correlation between droughts and El Niño mechanisms. This is reinforced by the fact that 8 of the 12 droughts that occurred in Southern Africa coincide with the years of the El Niño phenomenon.
By studying the correlation between precipitation and ENSO Nicholson and Kim (1997) asserted that the ENSO mechanism is causing precipitation and on the continent. Moreover, it is at the origin of numerous anomalies in Equatorial and Southern Africa. They also indicate that precipitation is negatively correlated with El Niño in the southern part of Africa. This correlation study was confirmed by Phillips et al. (1998), which showed that during the El Niño Southern Oscillation passage precipitation decreased, thus affecting agricultural production in Zimbabwe.
In southern Africa, droughts occur mostly during the warm El Niño Southern Oscillation Phillips et al. (1998). However, the El Niño mechanism is not the only factor causing drought in Southern Africa, as was the case in the 1970-88 period, there are other explanatory factors (Manatsa et al., 2008; Collier et al., 2008a). For example, droughts from 1950 to 1969 originate in the oceanic and atmospheric anomalies registered in the region.
In analyzing droughts from 1950 to 1988, Richard et al. (2001) argued that the droughts of 1970-1988 were severe and differed from those of 1950 to 1969 and that El Niño was not responsible for the 1925-1926 and 1997-1998 droughts in Southern Africa. As previously discussed, El Niño Southern Oscillation has two phases, one cold and the other warm. In the preceding paragraphs, it is clear that the hot phase that tends to affect Southern Africa. The main cause of drought (2010-2011) in East Africa was the Niña Dutra et al. (2013), which is the cold phase of the mechanism.
In trying to identify the cause of drought, Lott et al. (2013) found that human activities had no effect on short rains and that this drought was exclusively the work of the Niña in the Pacific (Haile, 2005; Tierney et al., 2013). The deficit in precipitation in West Africa is due to the warming of the Indian Ocean and the rise in greenhouse gas and aerosol emissions after the Second World War (Desboeufs et al., 2010; Funk et al., 2008; Williams and Funk, 2011).

Table of contents :

Introduction générale 
0.1 Mise en évidence du changement climatique
0.2 Changement climatique et secteurs agricoles
0.2.1 Effets du changement climatique sur l’agriculture
0.2.2 Répercussions sur la sécurité alimentaire
0.3 Plan de la thèse
1 Climate change and variability in countries and agroeological zones of the Sahel 
1.1 Introduction
1.2 Climate change: definitions, measures and models
1.2.1 Definitions
1.2.2 Measures
1.2.3 Models
1.3 Climate change and economic activities
1.3.1 Causes of climate change
1.3.2 The economic consequences of climate change
1.3.3 The channels of transmission of climate change
1.4 Study area and data
1.4.1 Study area
1.4.2 Data
1.5 Methodology
1.5.1 Pooled model with country effects
1.5.2 Structural change model
1.6 Application on different countries
1.6.1 Pooled model with heterogeneous coefficients
1.6.2 Temperature
1.6.3 Precipitation
1.7 Application on agroecological zones
1.8 Conclusion
1.9 Appendix
2 Climate and Agriculture : Empirical evidence for Countries and Agroecological Zones of the Sahel 
2.1 Introduction
2.2 Literature review
2.3 Econometric specification
2.4 Data
2.4.1 Perimeter
2.4.2 Production variables
2.4.3 Climatic variables
2.4.4 Other control variables
2.5 Results
2.5.1 Results at the country level
2.5.2 Results at the agroecological zone level
2.6 Conclusion
2.7 Appendix
3 Climate change and food security: a multidimensional analysis in the Sahel for the period 2000-2016 
3.1 Introduction
3.2 Literature review
3.2.1 Definitions of concepts around food security
3.2.2 Measures of food security
3.2.3 Causes of food insecurity
3.2.4 Economic consequences of food insecurity
3.2.5 Climate change and food security
3.3 Methodology and application: food security in Sahel
3.3.1 Construction of indices of the four dimensions of food security
3.3.2 Econometric specification
3.4 Data
3.4.1 Study area
3.4.2 Food security indicators
3.4.3 Variables of interest
3.4.4 Socioeconomic variables
3.5 Results and discussion
3.6 Conclusion
3.7 Appendix
Conclusion générale 
Bibliography
Appendix
3.8 Scope of study
3.8.1 Main economic characteristics by country
3.8.2 General situation of Sahel countries
3.9 Agronomic and food characteristics of the Sahel countries
3.9.1 Burkina Faso
3.9.2 Chad
3.9.3 Djibouti
3.9.4 Ethiopia
3.9.5 Mali
3.9.6 Mauritania
3.9.7 Niger
3.9.8 Nigeria
3.9.9 Senegal
3.9.10 Somalia
3.9.11 Sudan
3.10 Crop needs and stress: maize, sorghum, rice, wheat

GET THE COMPLETE PROJECT

Related Posts