Determination of coffee crops water requirements located in the Araguari River basin in Brazil

AVANS University of Applied Sciences – Breda, The Netherlands Environmental Sciences for Sustainable Energy Technology

Bachelor of Sciences

Jonathan Rivera

Student number: 2062300

GRADUATION PROJECT – FINAL REPORT

April 1

2014 to August 15

2014

Determination of coffee crops water requirements

located in the Araguari River basin in Brazil

Project mentor: Pr. Dr. Hudson de Paula Carvalho

University supervisor: Jappe de Best

Universidade Federal de Uberlândia – Instituto de Ciências Agrárias Uberlândia (MG) – Brazil

Advice report

Project: Graduation project from April 1st 2014 to August 15th 2014

Version: Final

University: AVANS University of Applied Sciences, Breda, Netherlands

Education: _{Environmental Sciences for Sustainable Energy Technology (ESSET)} Work

placement:

Universidade Federal de Uberlândia (Federal University of Uberlândia) Instituto de Ciências Agrárias – Engenharia Ambiental (Institute of Agrarian Sciences – Environmental Engineering)

Acknowledgment

I take this opportunity to express a deep sense of gratitude to Pr. Dr. Hudson de Paula Carvalho, my project mentor, for his cordial support, valuable information and guidance, which helped in completing this task on time, through various stages. He has been very kind and patient.

Secondly, I am very thankful to the Federal University of Uberlândia (UFU) and especially the Institute of Agrarian Sciences for having permitted me to perform this interesting project. I do not forget UFU International office for the help in the journey organization. They also provided to foreign students Portuguese classes for free, twice a week, which allowed me to live in better conditions in this amazing country and enhance my stay and integration.

I am obliged to thank Marcia Batistela Moraes, technician at the Federal University of Uberlândia for her cooperation and help during the period of my assignment.

I also take this opportunity to express my profound gratitude and deep regards to my University supervisor Jappe de Best, from Avans University of Applied Sciences in Breda, for his exemplary guidance and monitoring throughout the project.

In addition, I want to give many thanks to Avans University of Applied Sciences and all the teachers for having permitted me to perform my studies abroad, in the Netherlands, which allowed me to improve my English and my environmental sciences skills, and to meet unforgettable people from all around the world who made these two years one of the best experiences of my life.

I have a special thought to Norranny Lima, student from the Environmental Engineering course, for her great help for my integration within the university of Uberlândia and with translations from Portuguese to English, at work or during my daily life in Brazil.

Finally, my final thanks and not the least goes to my parents and my brother, for their constant encouragements and moral support all along my studies and during the time of this project. I would never have been able to achieve what I have realized so far without them.

Summary

One of the most common technologies adopted by farmers, especially the ones who grow their crop in the savanna, is irrigation. However there is no consensus about the management of this irrigation, especially in relation to water volume and irrigation frequency. This leads to an overconsumption of the resource. This study has allowed an estimation of the coffee crops water requirements over time in the Araguari River watershed, and the part of this water requirement that has to be provided by irrigation.

For a better accuracy, the whole watershed of the Araguari River was divided into three areas (see figure 5). To evaluate the water requirement of the coffee crops within each area, daily potential evapotranspiration rates have been computed from January 1st 2012 until December 31st 2013. For this, weather data and FAO Penman-Monteith method have been used to calculate the reference evapotranspiration of the crops and, combined with the crop coefficients, the annual optimal water requirements have been deducted. Then, to evaluate the water amounts to provide with irrigation in the coffee crops of each area, the Thornthwaite and Mather method has been applied to assess the water deficiency, based on the soil water balance throughout the year. Lysimeters have also been used in this purpose to allow direct experimental measurements, but the results were unusable.

Taking the average of 2012 and 2013, the calculated optimal water requirement for coffee crops gave annual results of 1370, 1110 and 1210 millimeters, respectively for the areas 1, 2 and 3. A part is provided by rainfalls but it has been estimated that the annual water deficiency was of 230, 64 and 167 mm respectively for the areas 1,2 and 3. Results from one year to another gave comparable values and water deficiencies were identified in June, July, August and September in the areas 1 and 3, corresponding to the southern dry winter, and in July, August and September in the area 2, where the temperatures are a little bit lower and the rainfall amounts higher in comparison to the two other areas.

To conclude, the water deficiency values previously mentioned give the annual amount of water to provide by irrigation. It has been seen that this irrigation is necessary during the southern winter, between June and September for the areas 1 and 3 and between July and September for the area 2. Thereby, an average annual amount of 230, 64 and 167 millimeters in the areas 1, 2 and 3 respectively has to be distributed during these months of water deficiency. The rest of the year, irrigation is not necessary since rainfalls provide enough water to the coffee crops. Also, the study has shown the importance of using an irrigation system with the highest efficiency possible because in term of water losses, the differences between two systems can be huge at the basin scale. The drip irrigation presented the best results.

List of tables

Table 1: Evapotranspiration conversion factors ... 5

Table 2: Coffee crop area per municipality and total area of the coffee crops within the Araguari River basin. ... 11

Table 3: Coffee crops size for each sub-area of the project ... 13

Table 4: Values of coffee crop Kc adopted for the project ... 20

Table 5: Optimal water requirement of the coffee crops in the 3 areas over the years 2012 and 2013. ... 27

Table 6: Water deficiencies and irrigation needs in the three areas in 2012 and 2013 ... 35

Table 7: Annual irrigation needs ... 36

Table 8: Annual water withdrawal for the irrigation of coffee crops ... 37

List of figures

Figure 1: The different types of evapotranspiration and their influencing factors 7 Figure 2: Araguari River basin borders and municipalities (A); localization of Minas

Gerais state in Brazil (B) and localization of the basin in Minas Gerais state (C). 8 Figure 3: Paranaiba basin and its 9 sub-basins (A); localization of the Paranaiba basin in

Brazil (B). 9

Figure 4: Annual precipitations and temperatures in Uberlândia (Minas Gerais) 10

Figure 5: Map of the watershed division into 3 sub-basins 12

Figure 6:Representation of the lysimeters used for the project 22

Figure 7: Underground view of two load cells placed under the lysimeter 1 23 Figure 8: Comparison between ETP and rainfalls in 2012 and 2013 in the area 1 29

Figure 9: Comparison between ETP and rainfalls in 2012 and 2013 in the area 2 30

Figure 10: Comparison between ETP and rainfalls in 2012 and 2013 in the area 3 31

Figure 11: ETA results and comparison with ETP in 2012 and 2013 for the area 1 32

Figure 12: ETA results and comparison with ETP in 2012 and 2013 for the area 2 33

Figure 13: ETA results and comparison with ETP in 2012 and 2013 for the area 3 34

Figure 14: Picture of a drip irrigation system in a young coffee crop 41 Figure 15: Picture of a center pivot (sprinkler) in a coffee crop 42

List of acronyms

AWC: Available Water Capacity ET: Evapotranspiration

ET0: Reference evapotranspiration

ETA: Actual evapotranspiration

ETP: Potential evapotranspiration

FAO: Food and Agriculture Organization Ha = Hectare

IBGE: Instituto Brasileiro de Geografia e Estatistica (Brazilian Institute of Geography and

Statistics)

ICIAG: Instituto de Ciências Agrárias (Institute of Agrarian Sciences) Km = Kilometer

Mm = Millimeter M3 = cubic meter

UFU: Universidade Federal de Uberlândia (Federal University of Uberlândia) UWR: Usable Water Reserve

Table of contents

1.     INTRODUCTION  ...  1  

1.1  GENERAL  BACKGROUND  ...  1  

1.1.1  Work  placement  ...  1  

1.1.2  Reason  of  the  project  ...  1  

1.2  GOAL  ...  2  

1.3  BOUNDARIES  ...  2  

1.4  READING  GUIDE  ...  3  

2.   THEORETICAL  BACKGROUND  ...  4  

2.1  DETAILED  PROJECT  CONTEXT  ...  4  

2.2  EVAPOTRANSPIRATION  PROCESS  ...  4   2.2.1  Definition  ...  5   2.2.2  Evapotranspiration  units  ...  5   2.2.3  Influencing  factors  ...  6   2.3  EVAPOTRANSPIRATION  CONCEPTS  ...  6   2.3.1  Reference  evapotranspiration  ...  6   2.3.2  Potential  evapotranspiration  ...  7   2.3.3  Actual  evapotranspiration  ...  7   2.3.4  Synthesis  overview  ...  7   2.4  STUDY  AREA  ...  8  

2.4.1  Araguari  River  basin  description  ...  8  

2.4.2  Climate  characteristics  ...  9  

2.4.3  Coffee  crops  within  the  Araguari  River  basin  ...  10  

2.4.4  Specific  areas  of  the  project  ...  11  

3.   METHODOLOGY  ...  14  

3.1  APPROACH  OF  THE  PROJECT  ...  14  

3.2  REFERENCE  EVAPOTRANSPIRATION  METHOD  AND  CALCULATIONS  ...  14  

3.2.1  Method  ...  14  

3.2.2  Calculations  ...  15  

3.3  POTENTIAL  EVAPOTRANSPIRATION  METHOD  AND  CALCULATIONS  ...  19  

3.3.1  Method  ...  19  

3.3.2  Calculations  ...  19  

3.4  ACTUAL  EVAPOTRANSPIRATION  METHOD  AND  CALCULATIONS  ...  21  

3.4.1  Method  ...  21  

3.4.2  Calculations  ...  23  

3.4.3  Irrigation  requirements  determination  ...  26  

4.   RESULTS  AND  INTERPRETATION  ...  27  

4.1  COFFEE  CROPS  POTENTIAL  EVAPOTRANSPIRATION  ...  27  

4.2  CLIMATIC  WATER  BALANCE  ...  29  

4.2.1  Area  1  ...  29  

4.2.2  Area  2  ...  30  

4.2.3  Area  3  ...  30  

4.3  COFFEE  CROPS  ACTUAL  EVAPOTRANSPIRATION  ...  31  

4.3.1  Lysimeters  ...  31  

4.3.2  Thornthwaite  and  Matter  method  ...  32  

4.4.1  Irrigation  requirements  of  the  coffee  crops  ...  34  

4.4.2  Water  withdrawn  from  the  basin  ...  36  

5.   DISCUSSION  ...  38  

5.1  INFORMATION  GATHERING  ...  38  

5.2  METHODS  AND  RESULTS  ...  38  

6.   REDUCTION  OF  THE  WATER  CONSUMPTION  ...  41  

6.1  IRRIGATION  TECHNIQUES  ...  41  

6.2  EVAPOTRANSPIRATION  REDUCTION  ...  42  

6.3  AWARENESS  AND  EDUCATION  CAMPAIGNS  ...  43  

7.   CONCLUSION  AND  RECOMMENDATIONS  ...  44  

7.1  CONCLUSION  OF  THE  PROJECT  ...  44  

7.2  RECOMMENDATIONS  ...  45  

8.   FURTHER  RESEARCHES  ...  46  

8.1  SUSTAINABLE  WATER  USE  WITHIN  THE  ARAGUARI  RIVER  BASIN  ...  46  

8.1.1  Water  availability  and  water  demand  within  the  Araguari  River  basin  ...  46  

8.1.2  Water  footprint  and  sustainable  assessment  of  the  Araguari  River  basin  ...  46  

8.2  IRRIGATION  SCHEMES  ...  47  

1

1. Introduction

In this part of the final report, information concerned with general background, project goal and boundaries is presented.

1.1 General background

1.1.1 Work placement

The Institute of Agrarian Sciences (ICIAG), part of the Federal University of Uberlândia (UFU) started its research and teaching activities in 1986 with the creation of Agronomy courses at Umuarama Campus in Uberlândia. The main activities of this institute are focused on the areas of plant pathology, crop sciences and soil sciences aiming to research and identify solutions for agricultural problems in Minas Gerais (region in which Uberlandia is located) and provide new options in order to lead to an agriculture as sustainable as possible. In 2009, an environmental engineering area was created, followed one year after by the environmental engineering course. The graduation project that has to be performed is part of this area. Indeed, environmental engineering education within ICIAG-UFU is mainly focused on the research and development of solutions that enable the harmonization of human activities with the physical environment and ecosystems, especially in the areas of water, monitoring and evaluation of environmental impacts of industrial and agricultural sectors and the management and evaluation of natural resources, using the most current and most sustainable technology available [1].

1.1.2 Reason of the project

Human activities - especially in the agricultural area with the irrigation for instance - use huge amounts of water and also cause pollution [2]. The problem is that less than 1% of the fresh water in the whole world is available for direct human uses. Yet, water is drawn inconsiderately. The current demand is such that it encourages drawing water everywhere, including where it will never be replaced [3]. In addition to the scarcity of the resource, population growth and economic development continually reduce the amount of water available. In a perspective of sustainable development, water must be kept intact for future generations, especially because its over-consumption and its pollution create complex problems at public safety, quantitative, qualitative, and economic levels [3]. Solving the problem at the source by reducing the rate of consumption is the basis of the conservation of this resource.

Brazil is the biggest producer of coffee in the world. An important part of this production occurs in the region of Minas Gerais. Based on the climate in this region, irrigation is essential for the coffee production in the crops, especially during the southern winter, which is dry. As it has been explained before, main demands and consumption of fresh water are for irrigation uses.

2 In this case, for the irrigation of the coffee crop in the region where the project takes place, water is taken from Araguari River basin. The growth of big urban cities such as Uberlândia has generated great pressures on water resources within this basin. In addition to this constant increasing for the water demand in the region, the contamination of these water resources by domestic wastewater, industrial and urban drainage lead to reduce strongly the water availability in the Araguari River basin [4].

That is why currently, it appears fundamental to start to reduce water consumption in this region. In order to manage it and to plan the water uses, a Watershed Committee has been created in 1997, bringing together both government entities and civil society. Several projects have already been carried out to preserve water resources of this basin [4]. ICIAG, thanks to Pr. Dr. Hudson de Paula Carvalho (civil society member), is involved in some projects of the Watershed Committee. As irrigation occurs for about 40% of the total water demand of the basin [5], it is essential to check if reduction of water consumption can happen in this area. That is why, one of the projects where the Watershed Committee of the Araguari River is involved, is to adjust the irrigation of crops according to their optimal water requirements. Another reason of reducing water consumption in coffee crops is, in an economic point of view, to decrease the production price of the coffee in order to make the coffee more profitable for the producers and even more competitive in the world market.

1.2 Goal

The goal of this graduation project is to estimate the water requirements of coffee crops located in the Araguari river basin, estimate the part of this water requirement that has to be provided by irrigation and find possibilities to reduce crops water consumptions within the basin. This study will allow avoiding water overconsumption of coffee crops and enable the elaboration of future effective irrigation plans, leading to a more sustainable consumption of water resources.

1.3 Boundaries

In order to yield a project, those prescribed limits explained below should be precluded of the project.

-­‐ Time: water availability is different according to the time during a year and even from one year to another one. That is why the result for this project will be an “estimation” of the water required. The data of two years (2012 and 2013) will be collected which will have the advantage to be quite significant.

-­‐ Space: the estimation of the water required and the amount to supply will concern only the Araguari river basin (approximately 20 000 km2).

3

1.4 Reading guide

In this report, chapter 2 describes the theoretical background of the project, focusing on evapotranspiration concept and the study area. In chapter 3, methodology about the approaches used can be found. Somehow, these two parts are a familiarization with the project about how it will be fulfilled to achieve the goal. Chapter 4 shows the results and their interpretation and is followed by a discussion in chapter 5. Chapter 6 is focused on the ways to reduce the water consumption for coffee farmers. The conclusion and recommendations are found in chapter 7 whereas chapter 8 will talk about the further studies that can be performed to extend and deepen the present project. At the end of this report, references and the appendices can be found.

4

2. Theoretical background

Theoretical information about the crop field evapotranspiration and the area of the project is described in this chapter. Also, a detailed context of the project can be found in the beginning of the chapter.

2.1 Detailed project context

Coffee production is an important part of the Brazilian economy. As a consequence, ways to enhance the production are actively researched for decades. Its climate is characterized by a bad rain distribution over the year, leading to drought in the southern winter, which lasts from April until September. As a consequence, crop productivities are affected. To face this problem, producers tool up more and more with irrigation systems, which allow to improve substantially the production yields, leading to increase their turnovers.

In Brazil, since environmental issues have just started to be taken into consideration, there is still no regulation about the water consumption. Also, producers are mostly independent farmers, irrigating without real plans or standards. In addition, water prices are very low, which can lead to assume that huge amount of water are wasted every year for the irrigation in the coffee crops. The Watershed Committee of the Araguari River has been created to avoid future water conflicts in the Araguari River basin. The first objectives are to quantify the water consumption within the basin (water demand) in order to make plans to lower it. Concerning the agriculture area, three main purposes are specified: evaluating the annual overall water requirements of coffee crops and according to the result, the water to supply with irrigation will have to be determined. Afterwards, effective irrigation schemes will be set up in order to avoid the over consumption by farmers [6]. As a consequence, all the stakeholders will feel benefits. Indeed, in addition to the reduction of the water use, producers will spend less money for the irrigation, while insuring to their production optimal growing conditions, leading to increase coffee productivity and quality. The present project is enrolled in the first two objectives, concerning the coffee crops located in the Araguari River basin. It is realized in collaboration with the Institute of Agrarian Sciences of the Federal University of Uberlândia.

2.2 Evapotranspiration process

The crop water requirement is defined by the amount of water needed to compensate the evapotranspiration loss from the cropped field. Thus, to estimate how much water is required by coffee crops in the project area, the evapotranspiration of these crops has to be computed. This part introduces the evapotranspiration.

5

2.2.1 Definition

Evapotranspiration is the combination between two processes, evaporation and transpiration: -­‐ Evaporation is the process whereby liquid water is converted to water vapor

(vaporization) and removed from the evaporating surface (vapor removal). Water can evaporate from many surfaces, such as lakes, rivers, soils, wet vegetation etc.

-­‐ Transpiration consists of the vaporization of liquid water contained in plant tissues and the vapor removal to the atmosphere. Plants mostly lose their water through stomata, located in the leaf.

It is hard to distinguish between these two processes since they occur at the same time. That is why it is preferable to combine them and talk about evapotranspiration [7].

Water quantities stored in a plant and those used by its metabolism are negligible in comparison with those that the plant must absorb due to losses through transpiration. Indeed, almost all water absorbed by plants is lost by transpiration. Therefore, mainly the evapotranspiration regulates water needs of plants [8].

2.2.2 Evapotranspiration units

Evapotranspiration rate is generally expressed in millimeters per unit time (per hour, day, month, decade, year or even a whole growing period). 1 millimeter equals to 0.001 meter and one hectare is equal to 10,000 m2. Thereby, if the crop loses 1mm of water, it is equivalent to a loss of 10 m3 of water per hectare (ha). A table of evapotranspiration conversion factors can be found in the table 1 below [7]:

Table 1: Evapotranspiration conversion factors

Depth Volume per unit area Energy per unit area

mm / day m3 / ha / day MJ / m2 / day

1 mm / day 1 10 2.45

1 m3 _{/ ha / day 0.1 1 0.245}

6

2.2.3 Influencing factors

Many physical, biological and meteorological factors can affect the evapotranspiration:

-­‐ Weather parameters: solar radiation, humidity, air temperature, atmospheric pressure and wind.

-­‐ Crop and soil characteristics: crop type, stage of growth, crop height, crop density, root depth and ground cover. The soil type and its water content may also influence the evapotranspiration of a crop field.

Management practices of the crop field such as cultivation or irrigation methods may also affect the microclimate or crop and soil characteristics [7]. Thereby, they might have an influence on evapotranspiration rates and thus, they should be taken into consideration when the evapotranspiration is computed, in order to have results as correct as possible.

2.3 Evapotranspiration concepts

It exists three types of evapotranspiration, the reference evapotranspiration (ET0), the potential

evapotranspiration (ETP) and the actual evapotranspiration (ETA). For the project, they have to be

calculated and this part explains their concepts, focusing on crop fields.

2.3.1 Reference evapotranspiration

Evapotranspiration from a reference surface with abundant water is denoted as ET0 and is called

the reference evapotranspiration. A hypothetical grass reference crop is taken as the reference surface and has specific characteristics.

Calculating ET0 is the first step for determining the water requirement of a given crop. It allows

to evaluate the evaporative demand of the atmosphere, without taking into account the crop type, soil and crop characteristics and management practices. Indeed, as water is abundant and always available, the soil is not considered as an influencing factor in the water absorption by the reference crop. In fact, only weather parameters affect ET0 rates. In other words, reference

evapotranspiration rates give the evaporating power of the atmosphere at a specific time and location.

ET0 has been developed because it presents some interesting advantages. Actually, ET0 values

measured or calculated in different seasons or at different locations can be compared since they refer to the evapotranspiration from the same reference surface.

Allen et al (1994) characterized ET0 as being the evapotranspiration of a hypothetical grass with a

height of 0.12m, an aerodynamic surface resistance of 70 s/m and an albedo (reflection coefficient expressing the reflecting power of a surface) of 0.23 [9].

7

2.3.2 Potential evapotranspiration

Besides the reference evapotranspiration, another necessary parameter for the calculation of the water requirement and the water to supply for a crop is the determination of the potential evapotranspiration (ETP). According to Allen et al. (1998), it is the evapotranspiration of a given

crop that occurs in an environment where the crop is implanted with the following standard characteristics: grown in large fields under optimum soil water content and agronomic conditions [7]. With these conditions, the potential evapotranspiration that is going to be calculated for coffee crops will express the maximal evapotranspiration rates or in other words, the optimal water requirement of the crop at the location of the study.

The ETP is generally estimated with the use of the crop coefficient (kc), where the direct relation

of this factor with ET0 gives the value of ETP.

Although it is the most used, it exists two other names to talk about the potential evapotranspiration: maximal evapotranspiration or evapotranspiration under standard conditions.

2.3.3 Actual evapotranspiration

The actual evapotranspiration, denoted as ETA, expresses the real evapotranspiration of the crop

under its own conditions that are different from the standard ones. Indeed, when crops are cultivated in fields, the actual crop evapotranspiration can differ from ETP due to non-optimal

conditions [7]. Therefore, when ETA is lower than the potential evapotranspiration ETP, it means

that the conditions are not optimal for plant growth.

2.3.4 Synthesis overview

A synthesis of this part can be found in the figure 1, where the different types of evapotranspiration are related to their influencing factors [7]. In this figure, E corresponds to evaporation and T to transpiration. ET0, ETc and ETc adj. refer respectively to the reference

evapotranspiration, the potential evapotranspiration of the crop and the actual evapotranspiration.

8

2.4 Study area

2.4.1 Araguari River basin description

The study takes place in the Araguari River basin, located in the mesoregions (Brazilian departments) of Triângulo Mineiro and Alto Paranaíba, in the western part of the state of Minas Gerais (figure 2). This is a sub-basin of the Paranaíba basin (figure 3). The basin has a size of approximately 22,000km2, covers about 20 municipalities and has altitudes ranging from 465 m to 1,350 m [10].

Figure 2: Araguari River basin borders and municipalities (A); localization of Minas Gerais state in Brazil (B) and localization of the basin in Minas Gerais state (C).

9

Figure 3: Paranaiba basin and its 9 sub-basins (A); localization of the Paranaiba basin in Brazil (B).

2.4.2 Climate characteristics

In the region of the Araguari River basin, rainfalls reach an average of about 1500 mm per year. The weather condition is warm, with an average annual temperature of about 21°C. There are two main seasons throughout the year, a wet season between October and March and a dry one during the southern winter, between April and September [11]. This climate makes the irrigation compulsory in the coffee crops, especially during the southern winter because precipitations are very low (figure 4) [12]. In the region of the study, according to the Köppen–Geiger classification system, the climate is Aw (savannah climate with a dry winter and temperatures higher than 18°C every month of the year).

10

Figure 4: Annual precipitations and temperatures in Uberlândia (Minas Gerais) 2.4.3 Coffee crops within the Araguari River basin

As mentioned previously, coffee production in the region is very important. It represents one of the biggest strengths of the economy of the region. Indeed, Minas Gerais is the state where the coffee production is the largest of the country with almost half of the total Brazilian production, especially in the project area, which is occupied by the “Cerrado”, a tropical savannah eco-region in Brazil [13].

For the project study, it is essential to know the area occupied by coffee crops in the Araguari River basin. This is possible thanks to the data found in the Brazilian Institute of Geography and Statistics website (IBGE) and which are summarized in the following table (table 2) [14]:

11

Table 2: Coffee crop area per municipality and total area of the coffee crops within the Araguari River basin.

As it can be seen, coffee crops occupy almost 100,000 ha, meaning approximately 1,000km2_.

This represents about a twentieth of the total area of the basin (22,000km2) and shows the important implantation of coffee crops in this region. Also, the coffee crops area per municipality within the basin is important to know, as it will be seen further in this report, for the calculations.

2.4.4 Specific areas of the project

As it has been explained in the part 2.2.3 “Influencing factors”, evapotranspiration rates depend of some weather parameters. Those parameters (and especially air temperatures) affect the water demand from crops and are also directly related to the altitude. Even if the climate is quite the same in the whole area, due to the size of the basin relatively small, it exists some differences that can have influences. As it has been impossible to collect meteorological data for all municipalities of the Araguari River basin, it has been decided to divide the project area into

sub-Municipality

_{area/municipality/(Ha)}

Coffee/crop/

Araguari 10424 Araxá 2506 Campos/Altos 9092 Ibiá 3456 Indianópolis 2942 Irai/de/Minas 794 Nova/Ponte 274 Patrocínio 29909 Pedrinópolis 317 Perdizes 6412 Pratinha 1872 Rio/Paranaíba 12393 São/Roque/de/Minas 2566 Sacramento 3089 Santa/Juliana 121 Serra/do/Salitre 10741 Tapira 317 Tupaciguara 136 Uberaba 878 Ubêrlandia 511 Total/(in/ha) 99748 Total/(in/Km2) 987

12 areas likely having the same weather conditions and thus similar evapotranspiration rates for coffee crops.

According to the IBGE website, medium altitudes and medium annual temperatures for each municipality have been gathered [14]. It has been seen that medium altitudes and annual temperatures are closely related and allow in this case a division of the Araguari River basin into three different sub-areas, by combining municipalities with same annual temperatures and approximately same medium altitudes:

-­‐ Area 1: Uberlândia, Uberaba, Tupaciguara, Santa Juliana, Nova Ponte, Araguari and Indianopolis;

-­‐ Area 2: Araxa, Campos Altos, Ibia, Pratinha, Perdizes, Rio Paranaiba, Sacramento, Tapira and Sao Roque de Minas;

-­‐ Area 3: Patrocinio, Irai de Minas, Pedrinopolis and Serra do Salitre.

A representation of the map with the 3 areas is shown in the following figure (figure 5):

Figure 5: Map of the watershed division into 3 sub-basins

For each area, one of the municipalities has a meteorological station and thus, climatic data can be collected. They are respectively Uberlândia, Araxa and Patrocinio for areas 1, 2 and 3. These

13 three municipalities are thereby considered as representative of their own area and their meteorological data will be used for the other municipalities of their whole area 1, 2 or 3.

For the calculations of the water requirement for coffee crops respectively in the 3 areas, the size of coffee crops for every sub-basin has to be known. Using the data of the table 2 in the part 2.4.3 “Coffee crops within the Araguari River basin”, a new table has been created, showing coffee crops size for the three sub-areas of the study (table 3):

Table 3: Coffee crops size for each sub-area of the project

Coffee%crop%area%(Ha)

14

3. Methodology

In this chapter, the different steps to reach the project goal are described. Methods and calculations are also explained.

3.1 Approach of the project

In order to reach the project goal, several steps have to be followed, allowing the estimation of the water requirement of the coffee crops located in the Araguari River basin and the part of this water requirement that has to be supplied with irrigation.

First of all, the optimal water requirement of coffee crops has to be determined. It is done by calculating the reference evapotranspiration. This reference evapotranspiration will allow to calculate the potential evapotranspiration of the coffee crops. Indeed, ETP refers to the water

losses by coffee crops under excellent water conditions and it is therefore the optimal amount of water needed by the coffee crops to compensate these losses.

Over the year, rainfalls provide a part of these water requirements of the coffee crops. Thus, the actual evapotranspiration (without taking into account the actual irrigation) must be measured or calculated and compared to the potential evapotranspiration to identify water deficiencies over time. This corresponds to the times over the year where rainfalls and soil water are not sufficient to provide enough water for allowing the plants to reach their potential evapotranspiration synonym of optimal water requirement. It is important that the coffee crops have an actual evapotranspiration similar to the potential one, because it means that the water provided is optimal, leading to a better productivity.

Therefore, the amount of water deficiencies will correspond to the amount of water to supply with irrigation. Afterwards, considering the efficiencies of the irrigation systems used within the project area, the amount of water withdrawn from the Araguari River basin to compensate with irrigation the water deficiencies of the coffee crops, can be estimated.

3.2 Reference evapotranspiration method and calculations

3.2.1 Method

It exists several methods for the estimation of the reference evapotranspiration. However, most of them are not very accurate and generally, ET0 is measured using meteorological data, because

this option gives the most accurate results. Also, it is very difficult to measure all the data required to calculate precisely ET0 by field measurements.

Many researches have been carried out and have leaded to a large number of empirical or semi-empirical equations in order to assess ET0 from weather data. From these studies, some methods

were developed and their own performances were tested. According to an Expert Consultation that has been held in 1990, the FAO Penman-Monteith approach is, from this date, recognized as the most effective and the standard method to measure ET0. This last has been the one used for

15 the reference evapotranspiration rates calculations of this project. Nevertheless, this method is not easy to use since it requires a lot of weather measurements, which can make it too expensive according to the research purpose. As a consequence, some other methods have been developed. They are simplified methods leading to less accurate results but depending of the study, they can be predominantly chosen because the calculations are easier and require more affordable material and less time or they can compensate missing data from Penman-Monteith equation.

3.2.2 Calculations

The reference evapotranspiration (ET0) has been calculated in a daily scale using the

Penman-Monteith equation as recommended by the standard FAO method [7]. The climatic data required for the calculations of ET0 were measured in Uberlândia, Araxa and Patrocinio stations,

representing respectively the areas 1, 2 and 3 (see figure 5).

Where ET0 is the reference evapotranspiration [mm day-1]; Rn is the net radiation at the crop

surface [MJ m-2 day-1]; G is the soil heat flux density [MJ m-2 day-1]; T is the mean daily air temperature at 2 m height [°C]; U2 is the wind speed at 2 m height [m s-1]; es is the saturation

vapor pressure [kPa]; ea is the actual vapor pressure [kPa]; (es–ea) is the saturation vapor pressure

deficit [kPa]; Δ is the slope of saturation vapor pressure curve [kPa °C-1] and γ is the psychrometric constant [kPa °C-1].

This equation cannot be solved directly. The different parameters T, Rn, G, γ, U2, es, ea and Δ will

be calculated thanks to the following equations proposed by the FAO standard. • Air temperature determination

The average air temperature T (or also called Tmean) was calculated from values of maximum and

minimum air temperatures, given by the 3 meteorological station located in Uberlândia, Araxa and Patrocinio respectively for the sub-area 1, 2 and 3.

With Tmax representing the daily maximum air temperature (°C) and Tmin the daily minimum air

16 • Air humidity determination

To obtain the saturation vapor pressure, it was used the equation given by Tetens: 𝑒_! = 0.6108  ×  10!"#.!!!!.!!

Then, actual vapor pressure can be determined from the difference between the dry and wet bulb temperatures, the so-called wet bulb depression. The relationship is expressed by the following equation:

e

= e° (T

) – γ . (T

- T

)

Where ea is the actual vapor pressure [kPa]; e°(Twet) is the saturation vapor pressure at wet bulb

temperature [kPa]; γ is the psychrometric constant of the instrument [kPa °C-1]; Tdry-Twet is the

wet bulb depression (with Tdry the dry bulb and Twet the wet bulb temperature [°C]).

For the determination of e°(Twet), Tetens equation may be used, replacing T by Twet.

Also, the psychrometric constant was given by the values of atmospheric pressures measured in a daily scale in the three meteorological stations of the study:

γ = A × P

Where A is a coefficient depending on the type of ventilation of the wet bulb [°C-1], and P is the atmospheric pressure [kPa]. For this study, A is equal to 0.000800 °C-1 because it is a natural ventilated psychrometer.

Finally, the slope of saturation vapor pressure curve was obtained from values of average air temperatures and saturation vapor pressure:

𝛥 = 4098  ×   𝑒! (𝑇 + 237.3)! • Wind speed determination

Wind speed was measured in the 3 stations of the study, but at a height of 10 meters. According to the FAO standard for the Penman-Monteith equation, the wind speed has to be measured at a height of 2 meters. To adjust this, a logarithmic wind speed profile may be used, in order to estimate the wind speed at a height of 2 meters:

17 Where U2 is the wind speed at 2 m above ground surface [m s-1]; Uz is the measured wind speed

at z meter(s) above ground surface [m s-1] and z is the height of measurement above ground surface [10 meters for the 3 stations].

• Radiation determination

The net radiation (Rn) was calculated as being the difference between the incoming net shortwave

radiation (Rns) and the outgoing net longwave radiation (Rnl)

R

= R

– R

Where Rn is the net radiation at the crop surface [MJ m-2 day-1]; Rns is the net ingoing shortwave

radiation [MJ m-2 day-1] and lastly, Rnl is the net outgoing longwave radiation [MJ m-2 day-1].

Rns was estimated based on grass albedo and the incoming solar radiation:

R

= (1 - α) × R

Where α is the albedo or canopy reflection coefficient, which is equal to 0.23 for the hypothetical grass reference crop [dimensionless] and Rs is the incoming solar radiation [MJ m-2 day-1].

The incoming solar radiation (Rs) was estimated using the Angström-Prescott equation, which

relates solar radiation to extraterrestrial radiation and relative sunshine duration:

Where n is the actual duration of sunshine [hour/day]; N is the maximum possible duration of sunshine or daylight hours [hour/day]; n/N is the relative sunshine duration [-]; Ra is the

extraterrestrial radiation [MJ m-2 day-1]; as is the regression constant, expressing the fraction of

extraterrestrial radiation reaching the earth on overcast days [dimensionless] and bs is the

coefficient of Angström-Prescott equation [bs = 0.54].

The coefficient “as” has been calculated using the following formula proposed by Glover and Mc

Culloch:

a

= 0.29 × cos φ

18 Daily values of the extraterrestrial radiation were obtained from the following equation:

𝑅_! = 37.6  ×  (𝑑 𝐷)!  ×  [  

𝜋

180° ×  ℎ𝑛  × sin 𝜑 × sin 𝛿 +  cos 𝜑 × cos 𝛿  × sin ℎ𝑛  ] Where (!_!)! _{is the relative distance Earth-Sun; hn is the sunset hour angle [in °] and δ is the solar} declination [in °].

The relative distance Earth-Sun  (!_!)!_{, the solar declination “δ” and the sunset hour angle “hn” are} given respectively by the three formulas below:

(𝑑

𝐷)! = 1 + 0.033  × cos  (𝐷𝑁  ×   360 365  ) Where DN refers to the day number in the year [between 1 and 365].

𝛿 =  23.45  × sin  [  360

 365  ×   𝐷𝑁 − 80 ] ℎ𝑛 = arccos  [−𝑡𝑔   𝜑  ×  𝑡𝑔   𝛿 ]

The daily photoperiod was computed from sunset hour angle (hn) values: 𝑁 = 2  ×  ℎ𝑛

15°

Following the previous equations, Rns can now be calculated. Regarding the first equation Rn =

Rns – Rnl, it remains to find Rnl. It is given in the equation hereafter:

Where σ is the Stefan-Boltzmann constant [4.903 10-9 MJ K-4 m-2 day-1]; Tmax,K is the maximum

absolute temperature during the 24-hour period [K (= °C +273.16)]; Tmin,K is the minimum

absolute temperature during the 24-hour period [K]; ea is the actual vapor pressure [kPa] already

calculated; Rs is already calculated and Rso is the clear-sky radiation [MJ m-2 day-1].

The estimation of the clear-sky radiation Rso can be found in the equation given below:

R

= (0.75 + 2.l0

z) × R

19 Finally, the last parameter to calculate is G, the soil heat flux. It has been done following the equation proposed by Wright and Jensen

G = 0.38 × (T – T

)

Where T-3d refers to the average air temperatures of the 3 last days [°C].

Following all these equations, the reference evapotranspiration rates (ET0) can be computed in a

daily scale in the 3 sub-areas of the study. The results are then classified and organized in an Excel table.

3.3 Potential evapotranspiration method and calculations

3.3.1 Method

Differences in evapotranspiration between reference grass surface and a crop field under the same (optimal) conditions can be expressed by a crop coefficient. ETP differs from ET0 because the

crop type and height, canopy properties and aerodynamic resistance of the crop are different from the ones of grass. Also, the coefficient kc varies over a year according to the crop growth stage. In order to estimate the potential evapotranspiration under optimal conditions, the coefficient kc of the crop studied has to be determined [15].

3.3.2 Calculations

After having calculated reference evapotranspiration rates, the second step is the estimation of ETP, the evapotranspiration under standard conditions or maximal evapotranspiration. It is

determined by the following formula:

𝑬𝑻𝑷 =   𝑬𝑻𝟎  ×  𝑲𝒄

Where ETP is the crop evapotranspiration under standard conditions [mm day-1]; Kc is the crop

coefficient [dimensionless] and ET0 is the reference crop evapotranspiration [mm day-1].

Weather conditions are incorporated in the reference evapotranspiration calculations. As a consequence, Kc values vary with crop characteristics and not according to the specific climate. This can enable the transfer of standard Kc values between locations for a similar crop type [7]. Thereby, for this study, it has been used Kc values measured over a previous project realized by Camargo and Pereira (1990), about coffee plants in Brazil [16]. These values can be found in the table 4, below:

20

Table 4: Values of coffee crop Kc adopted for the project

Kc gives monthly values whereas ET0 is measured everyday of the year. Then, kc value of a

given month has to be repeated for each day of this month in order to determine the daily ETP.

Coffee plants are now well adapted with the planting system in mechanized hedgerows, characterized by a thickening of the plants in the same row, and a wider spacing between rows of planting, providing optimal space for traffic of agricultural machinery. Currently, this system is the predominant one in Brazil, which has replaced the previous system, where the spacing between plants was considerably larger. Moreover, there is no predominance in the spacing between plants or the inter-rows, which influences the number of plants per area. However, it has been adopted for this project the following characteristics because they are considered as the most common in the study area:

-­‐ The distance between plants in a same row is equal to 0.7 meters; -­‐ The distance between rows equals 3.5 meters.

In this arrangement, available area for each plant is of 2.45 m2 and achieves then a population of 4081 plants per hectare.

Another factor that was considered is the crop coverage rate over the ground because for the irrigation method used, currently, there is a convergence for two types of systems:

-­‐ Localized, with the use of drip irrigation;

-­‐ Sprinkler, using center pivot equipped with a system that applies water in greater quantities and at low pressure on plants (LEPA).

For both methods, inter-rows are not irrigated implying some changes in the processes of crop water loss and thus, affecting the crop evapotranspiration. Therefore, some adjustments were adopted and are shown in the following formulas [15]:

𝐸𝑇_! = 𝐸𝑇_!  ×  𝑘_!  ×  𝑘_! 𝑘! = 𝑇! +  

1

2  ×  (1 −  𝑇!)

Month

Kc value

Month

Kc value

January 0.89 July 0.73 February 0.87 August 0.73 March 0.91 September 0.74 April 0.79 October 0.89 May 0.73 November 0.90 June 0.73 December 0.95

21 Where ETP is the adjusted potential coffee crop evapotranspiration [mm day-1]; kr is a coefficient

that depends on the crop coverage rate over the ground [dimensionless] and Tc is the crop

coverage rate over the ground [dimensionless].

The crop coverage rate over the ground was obtained by the relation between the projected area of the crop canopy and the ground area available for crop development. It was considered in this work that each coffee plant has a projected canopy area over the ground of 1.26 m2, which is very common for mature coffee plants and an available area for each plant of 2.45 m2, as explained previously in this chapter. Thus, Tc and the coefficient kr can be calculated:

𝑇_! =  1.26

2.45= 0.51 𝑘_! = 0.51 +  1

2  ×   1 − 0.51 =  0.755

As ETP represents the potential evapotranspiration (or maximal ET) of coffee crops under

optimal conditions (no water shortage), once calculated, it gives the optimal water requirement of crops in the areas of the project.

3.4 Actual evapotranspiration method and calculations

3.4.1 Method

The actual crop evapotranspiration shows the evapotranspiration of a given crop in the real conditions of its cultivation as explained previously. Basically, it exists some methods to measure or estimate it, directly or indirectly. Unfortunately, for most of them, some parameters could not be calculated because of a lack of data or material, especially for soil and crop characteristics. That is why for the purpose of this project, only two methods were available: experimentally using lysimeters, which allow direct measurements and a theoretical method, from Thornthwaite and Matter.

• Approach 1: lysimeters

A lysimeter is used to study and measure water evolutions in a natural soil, agricultural crop, forest etc. Although it is difficult and expensive to implement, lysimeters are sometimes used for specific researches such as the actual evapotranspiration determination in agricultural crops, in order to develop effective irrigation schemes. It exists two types of lysimeters [17]:

-­‐ Weighing lysimeters -­‐ Non-weighing lysimeters

22 For this project, weighing lysimeters were used. The principle is to fulfill an isolated tank with soil. By recording, precipitations (and eventually the actual irrigation), the drainage and the changes in soil water content, the actual evapotranspiration rates can be computed. When they are well calibrated and well monitored, lysimeters allow the direct determination of the actual evapotranspiration with a very good accuracy, while taking into account the soil water balance over time [17].

The lysimeters are installed in the experimental farm named Gloria, belonging to UFU. They are located in Uberlândia, in the following geographic coordinates: 18° 58’ 52’’ (south latitude) and 48° 12’ 24’’ (west longitude), with an altitude of 912 meters [18].

In this area are installed three weighing lysimeters, in a rectangular/square shape, with the following characteristics:

-­‐ Lysimeter 1: 0.990m (width); 1.358m (length) and 0.7m (depth) -­‐ Lysimeter 2: 1.355m (width); 1.352m (length) and 1.4m (depth) -­‐ Lysimeter 3: 1.795m (width and length) and 2.0m (depth). A scheme of the three lysimeters is represented in the figure 6 [18]:

Figure 6:Representation of the lysimeters used for the project

• Approach 2: Theoretical method

Besides the use of lysimeters giving direct measurements of ETA, a theoretical approach can be

applied in the conditions of the project for the estimation of actual coffee crops evapotranspiration rates in the study area. It can be done by estimating the soil water balance over time, with the method developed by Thornthwaite and Matter [19]. It allows the determination of ETA without the need for direct measurements of soil and crop conditions.

For its elaboration, it is necessary first of all to know three parameters:

-­‐ The maximum water storage in the soil (AWC - Available Water Capacity); -­‐ Total rainfall;

23 -­‐ Potential evapotranspiration rates (so called maximal evapotranspiration of the studied

crop).

The two last parameters have to be available at a daily or monthly scale.

3.4.2 Calculations

• Experimental measurements from lysimeters

The lysimeters can allow the measurement of the actual evapotranspiration ETA with the

following equation [20]:

ET

= R – D – ΔS

Where: R is the rainfall amounts [mm]; D is the drainage [mm] and ΔS is the difference in soil mass, representing actually the variations in soil water content within the lysimeter [mm].

These lysimeters measure D and ΔS every 5 minutes. The results are stored in a datalogger located close by and can be downloaded at any time via computer with a specific software. The rainfalls are measured at a daily scale in the meteorological station located in Uberlândia.

To measure ΔS, the weight fluctuations of the lysimeters (changes in soil water content every five minutes), it is used load cells placed underground, under the lysimeters (figure 7). For each lysimeter, 4 load cells are used. The models used are LC501-2K, LC501-3K and LC501-5K (OMEGA Engineering ®) for the lysimeters 1, 2 and 3 respectively [18].

Figure 7: Underground view of two load cells placed under the lysimeter 1

A load cell is a transducer, which converts force/weight variations into a measurable electrical output. Every five minutes, measures are sent to the datalogger. When data is downloaded, an equation is applied, allowing to convert results in millivolt into millimeter of water. This equation

2 out of 4 load cells

24 is determined during the calibration of the lysimeters, made in March 2013. Indeed, the good calibration of the lysimeters allowed to build a linear regression curve thanks to the proportionality between the voltages measured and the height of water tested [18]. Each lysimeters has its own regression line:

-­‐ Lysimeter 1: Water height (mm) = 919.7 * Tension – 1188.2 -­‐ Lysimeter 2: Water height (mm) = 1028.78 * Tension – 1919.79 -­‐ Lysimeter 3: Water height (mm) = 922.77 * Tension – 3042.53

Thereby, once the data downloaded, the measures are directly converted to millimeter in an excel file and the difference in soil water content is equal to:

ΔS = M

- M

Where: Minitial is the initial water height of the lysimeter at t0 [in mm] and Mfinal is the water height

of the lysimeter at t = t0 + 5 minutes [in mm].

For the purpose of the project, irrigation is not taken into account in the calculations in order to make a comparison between the water requirements under optimal conditions and the actual behavior of plants without irrigation. Thereby, if a water deficit is observed, the water amount to supply with irrigation will be deducted. Actually, the lysimeters 1 and 2 do not receive the irrigation applied for the coffee crops of Gloria farm contrary to the lysimeter 3. Initially, this has been done in order to meet two objectives. Indeed, the calculation of the potential evapotranspiration made in parallel, gives the optimal water requirement of coffee crops. Then, measurements of the actual evapotranspiration by the lysimeters 1 and 2 where the irrigation is not taken into account will allow the evaluation of the optimum water amount to supply with irrigation, since only the water furnished by the rain is measured in the calculations. On the other hand, the ETA results with the lysimeter 3 where irrigation is included will allow a comparison

between the optimum water amount to supply with irrigation and the actual irrigation amount applied for coffee crops at the study location.

As a consequence, the equation for the ETA measurements established for the lysimeters 1 and 2

is different for the lysimeter 3. It becomes:

ET

= R + I – D – ΔS

Where I is the amount of water provided by irrigation [mm].

Unfortunately, this irrigation amount was unable to be measured and so, ETA cannot be

calculated regarding this equation. The lysimeter 3 is thus unusable for this project and only results from the lysimeters 1 and 2 will be assessed.

25 Thus, a comparison between the optimal irrigation and the actual one will not be possible. As there is no data available currently about irrigation since there is no plan or standards, only the optimal water to supply with irrigation will be assessed.

• Theoretical calculations using the Thornthwaite and Matter method

The previous method using lysimeters allows to determinate directly actual evapotranspiration rates over time of the coffee crops, because the soil water balance (soil water content over time) is taken into account in the calculations. ETA can also be calculated theoretically using the

Thornthwaite and Matter approach. This is done when direct measurements cannot be realized (experimentally with lysimeters for instance) or to make a comparison with experimental measurements and prove or refute their accuracy.

The Thornthwaite and Matter method allows to calculate actual evapotranspiration rates by comparing the potential evapotranspiration ETP with the rainfall. But this comparison is not

sufficient. Indeed, the soil water content over time must also be taken into consideration. For instance, if for a certain day, the potential evapotranspiration is of 3 mm and the rainfall equals to 20 mm, then precipitations are enough to compensate the water losses by evapotranspiration. The next day, if the potential evapotranspiration is equal to 2 mm for example, but there is no rain, one may considerate that water must be added in order to compensate losses by ET. This is wrong because there is still water available for plants in the soil, thanks to the large amount of rain from the last day. Therefore, this method tries to estimate the soil water content over time (so called soil water balance) in order to determine the irrigation water needs by plants in a way as accurate as possible, by specifying the water deficiency.

Thus, before starting the calculations, an important soil characteristic has to be known, the available water capacity (AWC). It represents the maximum amount of water available for plants that a given soil can store. According to a previous research made by Carvalho (2008), AWC is equal to 100 mm for coffee crops in the study area [15]. These 100 mm are the usable water reserve of the soil (UWR), meaning the amount of water that the soil can absorb and release to the plants. In other words, it is the difference between moisture at the AWC and moisture at the permanent wilting point PWP. When the soil reaches the AWC (100 mm), it is saturated because all the pores are filled with water. Then, if water is added (by rainfall or irrigation), this surplus will either go directly to the depths of the soil by percolation, (named gravitational water because it is not retained by capillary forces), or either run off. If water is not added when the soil is saturated, plants will pump the UWR to compensate losses by ET. Afterwards, if it does not rain or if irrigation does not occur, the PWP is reached and will lead progressively to the plants death [21].

The Thornthwaite and Matter method considers three cases. First of all, if precipitations (P) exceed the potential evapotranspiration, the amount of water corresponding to P - ETP is stored in

26 The coffee crops growth conditions are optimal and therefore, the actual evapotranspiration is equal to the potential evapotranspiration (maximal ET under optimal conditions).

In the case where P = ETP, ETP = ETA. Once again, enough water is available for optimal coffee

crop growth, but in this case the usable water reserves of the soil stay the same than the last day or month (according to the scale of the measurements), since all the water added is used by plants to compensate the ET.

Finally, when P < ETP, plants will use the water provided by rainfall plus the usable water

reserve. As a consequence, ETA = P + a part or all the UWR until its depletion. When the usable

water reserve is depleted and ETP > ETA, the difference ETP – ETA corresponds to the water

deficiency [21].

3.4.3 Irrigation requirements determination

Both methods, lysimeters and Thornthwaite-Monteith, are going to inform about actual evapotranspiration rates of the coffee crops. Afterwards, ETP and ETA values will be compared

for both methods. Indeed, as explained previously, for an optimal coffee production rate, ETA

should be equal to the potential evapotranspiration. So, the difference ETP – ETA will be checked

up. When ETP and ETA are similar, it will mean that the water provided by rainfall will be

sufficient to satisfy the water needs of the coffee plants and then, irrigation will not be necessary. On the other hand, when ETA values are lower than the ETP ones, it will reflect a lack of water

(water deficiency) and mean that irrigation is strongly recommended. In this case, the amount of irrigation to supply is equal to the water deficiency.

Finally, the water deficiency in millimeter, calculated in the three municipalities where are located the meteorological stations (one for each sub-area of the Araguari River watershed) will be translated into m3_{/ha and extrapolated to the municipalities of their own area according to the}

total coffee crop sizes of the sub-basins presented in the part 2.4.4 “Specific areas of the project”. This will represent the annual volume of water to supply with irrigation for every area. The sum of the results from the 3 areas will give the total annual volume of water to supply with irrigation in the whole basin. Knowing this last amount, the total amount of water withdrawn to the Araguari river basin for irrigation needs of coffee crops will be determined. It will represent the optimal amount that should be used in order to set up future efficient irrigation schemes and avoid current potential overconsumption. This amount of water to withdraw from the basin will be equal to the total annual volume of water to supply with irrigation in the whole basin multiplied by the efficiency of the irrigation used.

27

4. Results and interpretation

This chapter presents the results of the project about the different evapotranspiration rates calculated, coffee crops water requirements and irrigation needs. An interpretation of these results is also done. In order to ensure clarity, the daily results were translated into monthly results, by adding up ETP values of each day of a month.

4.1 Coffee crops potential evapotranspiration

First of all, reference evapotranspiration rates have been calculated using the FAO Penman-Monteith standard method, in order to evaluate the evaporative demand of the atmosphere over time thanks to a reference surface (hypothetical grass with standardized characteristics). Daily results of ET0 values allowed to calculate the coffee crops potential evapotranspiration rates by

applying the formula given in the chapter 3.2 “Potential evapotranspiration calculations”:

ET

= ET

x k

x k

= ET

x k

x 0.755

ETP calculations are going to show the optimal water requirements of the coffee crops over the

years 2012 and 2013 in the three sub-areas of the project. The results are shown in the table 5 below, where the results are expressed in millimeter:

Table 5: Optimal water requirement of the coffee crops in the 3 areas over the years 2012 and 2013.

Area 1

Area 2

Area 3

Month 2012 2013 2012 2013 2012 2013 January 158 151 136 122 137 126 February 152 145 122 109 122 112 March 126 127 124 99 126 106 April 109 112 102 80 96 81 May 91 95 81 68 82 72 June 87 76 54 51 73 65 July 95 53 53 51 79 53 August 99 55 63 61 86 47 September 122 92 87 73 120 88 October 139 123 101 96 128 102 November 126 121 117 111 132 113 December 147 138 129 125 141 132 Total (in mm/year) 1451 1288 1169 1046 1322 1097

According to these results, this table shows that the water requirement for coffee crops in the Araguari River basin is approximately between 1000 and 1500 mm per year, or between 10,000 and 15,000 m3_{/ha/year. In a general point of view, for each area, the results are comparable}

28 between 2012 and 2013, even if the monthly potential evapotranspiration rates are a little bit lower in 2013 than in 2012, due to differences in climatic conditions between the two years. This similarity between 2012 and 2013 for each area might prove the accuracy of the results. Also, the research made by Bueno (2012) showed than generally, coffee crops water needs are in a range between 800 - 1200 mm according to the location, which is quite consistent with the results found during the present project, within the Araguari River watershed [22]. For the area 1, this water requirement is a little bit higher than 1200 millimeters in 2012 and 2013 and for the area 3, it is higher in 2012. This can be explained by the kc values taken for the project and which are monthly averages, making them less accurate.

However, regarding the monthly values, it can be perceived that the potential evapotranspiration is higher during the southern summer where it is warm and humid than during the south winter (April-September) where it is a little bit colder but dry. This can appear contradictory since the humidity is much lower in the dry season and the strong rainfall amounts in the summer may suggest that the radiation is lower in the summer. These two parameters lead to an average evapotranspiration lower in the wet season (October-March) than for the dry one. Nevertheless, a research made about the evapotranspiration in the Brazilian tropical savannah shows the same ET fluctuations over the year, with the same climatic characteristics than for the Araguari River watershed [23]. The authors supposed that some climatic characteristics such as air temperatures, net radiation and duration of the days are higher in the wet season and explain then, the higher ET calculated in this period of the year. Indeed, even if the climate is very humid during the southern summer and is marked by strong rainfalls, it has been shown in a study made by Gonçalves dos Santos (2008) that in this season, the rainfall is characterized by thunderstorms with a huge amount of rain during one or two hours. The rest of the day is sunny and with a clear sky. This demonstrates why net radiations are higher in this period than during the dry winter. Also even if the temperatures stay warm during the winter, the nights are way colder leading to mean temperatures significantly lower than during the summer [24]. Correlated with the duration of the days more important during the summer, all these characteristics explain why the ETP rates are

the highest in the wet season, even with higher humidity than during the dry season from April to September.

Finally, it can be seen that the potential evapotranspiration rates calculated were higher in the area 1 than for the two other ones. If the average between 2012 and 2013 is made, the optimal water requirement for coffee crops was of 1370 mm in the area 1, about 1100 mm in the area 2 and 1200 mm in the area 3. This can be explained by the difference in altitudes. Indeed, higher the altitude, lower the temperature leading to a lower potential evapotranspiration. This is proven by the weather data provided by the meteorological stations of the three areas where the mean annual temperatures are of 21.9°C, 20.4°C and 20.8°C respectively for the areas 1, 2 and 3 [25].