1. Introduction
Most scientists agree that global climate change is occurring at an alarming rate and that these changes are likely to impact water use in horticulture, agronomy, and natural ecosystems. In some locations, climate change can potentially increase agricultural production, but it is generally believed that widespread detrimental impacts on agricultural production are more likely in much of the world. Currently, the global climate change trend is for increasing air temperature, mainly at night and during winter, and more near the poles than in lower latitudes. Because of global warming, energy storage in water has increased dramatically (much more than in the air), and higher water temperatures has led to rising sea level due to water expansion and additional heat storage mainly in the oceans. In addition, increasing air and water temperature enhances evaporation, and higher air temperature increases the saturation water vapor pressure, i.e., the amount of water vapor that is held in the air at saturation. These atmospheric changes can impact plant growth, production, and water usage, and this chapter presents ideas on the possible impact of higher temperature, humidity, and CO2 concentration on the evapotranspiration (ET) of horticultural crops.
At this time, the main cause for global climate change is the increasing concentration of CO2. According to the “Keeling curve,” the CO2 concentration recorded at the Mauna Loa Observatory in Hawaii (USA) increased from about 315 to 395 ppm from 1958 to 2013, and recently passed 400 ppm. This increase has serious implications for global climate change and its impact on nature and horticulture.
The greenhouse gases (GHGs) that contribute most to global climate change from the worst to least worse are as follows: (1) H2O—water vapor; (2) CO2—carbon dioxide; (3) NH4—methane; (4) N2O—nitrous oxide; (5) O3—ozone. While H2O is actually a more effective GHG than CO2, the atmospheric concentration is spatially and temporally variable and it is not increasing rapidly at this time. However, as the oceans warm, higher concentrations of atmospheric H2O are likely, and the higher levels can greatly contribute to warming. Carbon dioxide is a less effective GHG than H2O, but the more evenly distributed global concentration is increasing steadily, and it is currently the GHG causing the most rapid global temperature rise. Methane is a concern for the future because large amounts of methane are stored in the ocean and in permafrost. Scientists believe that thawing the permafrost and warming the oceans might release this stored NH4 and cause a rapid temperature increase. Nitrous oxide is used as an aerosol propellant, anesthetic, and an oxidizer in rockets and engine fuel. Ozone is not evenly distributed over the Earth’s surface, but man-made O3 does contribute to atmospheric warming in urban areas.
GHGs mostly reside in the troposphere, i.e., the lowest level of the atmosphere, where they intercept long waveband radiation (LWR) and raise atmospheric temperatures. The stratosphere, which is a stable layer of air above the troposphere, has considerably less greenhouse gas than the troposphere, and, interestingly, this upper atmospheric layer is cooling. If the global warming were caused by increasing solar output or reduced reflection of solar radiation, the troposphere and stratosphere would both likely have increasing temperatures. Consequently, the simultaneous troposphere warming and stratosphere cooling is a good indicator that the warming is caused by anthropogenic increases in GHG additions (mostly CO2) to the atmosphere. The melting of polar ice packs is a strong indication of global warming, and it will likely compound the warming because less solar radiation will be reflected back to space as the ice melts.
Climate change is likely to cause sea level rise and damage plant and animal life in the oceans and on land. Of course, changes in temperature and humidity could also lead to changes in general circulation of the atmosphere, greater frequency of storms and floods, and changes in the length and severity of drought. In general, climate scientists are projecting more frequent and severe storms and drought due to changes in atmospheric circulation [
1]. In terms of horticulture, the authors of [
2] and [
3] discuss several potential effects of climate change on horticulture. In some cases, increasing CO
2, temperature, humidity, and other greenhouse gases might be beneficial in regions where crop production is limited by cold temperatures due to (1) lower potential for frost damage, (2) faster growth, and (3) lengthened growing seasons. In addition, there is some benefit coming from CO
2 fertilization, which can enhance photosynthesis. On the other hand, climate change could negatively impact agriculture in regions where climate conditions are currently good for production. For example, climate change might (1) decrease chilling and inhibit bloom and fruit set in horticultural crops, (2) lead to high temperature and wind during bloom or ripening that could negatively impact fruit set or fruit quality, (3) increase ET rates that could lead to water deficits, and (4) increase problems with heat stress.
Some possible climate change impacts on agriculture include (1) droughts, (2) floods, (3) faster phenological development, (4) inadequate chilling requirements, (5) pollination affected by rainfall and other extreme events, (6) frost and chill damage, (7) the spread of new insects and diseases, and (8) lower or higher yield and quality due to warming and water relations during summer [
2,
3]. One example of warming impacts is decreased winter chilling, which leads to bad pollination, staggered bloom, reduced fruit set, and poor fruit quality. In low chill years, apricots and cherries can drop flower buds or not produce a crop when chilling is inadequate.
In some regions, winter fog is an important factor in achieving adequate chilling for some crops. Rainfall during bloom can inhibit bee pollination, and precipitation around harvest time can increase fruit and nut diseases, cause fruit cracking, and destroy a crop. In deciduous orchards, late spring and summer rainfall has a negative effect on fruit and nut quality and production. Clearly, there are a multitude of climate factors that can change to wreak havoc on the production of horticultural crops. For example, climate scientists are projecting more bimodal precipitation in California, with more precipitation in the spring and fall and less in the winter. However, California depends on water storage in the mountain snowpack, and higher snow lines in the mountains, especially in the north where the mountains are shorter, will have less snowpack storage and will result in less water delivery to agricultural land in the summer [
4]. Snowpack storage will affect the collection and distribution of irrigation water, but precipitation timing will also affect rainfall damage to crops in the spring and fall as well as changes in fog formation due to winters with higher temperatures and less rainfall.
While all of the aforementioned changes and impacts are important to consider, it is difficult to project changes in general circulation, storms, and drought with a high degree of accuracy in any particular region on Earth. However, the general impact of widespread higher temperatures, humidity, and increased CO2 on factors affecting crop production is possible. Following a short discussion on how rising CO2 concentrations cause global warming, this paper will present some ideas on the possible impact of higher temperature, humidity, and CO2 on ET.
2. Increasing CO2 and Global Warming
The impact of atmosphere CO
2 concentration on the greenhouse effect and the Earth’s surface temperature was first described by Svante August Arrhenius [
5]. Although climate science has advanced considerably since Arrhenius first made his estimates of increasing CO
2 concentration, its impact on global temperature is still quite accurate. Equations (1)–(3) are modified versions of the original equations from [
5]. Equation (1) provides an estimate of Earth’s emission temperature (
Te), which is the mean surface temperature it would have if there were no atmosphere:
where
RSC = 1631 W·m
−2 is the solar constant,
αp ≈ 0.30 is the albedo (reflection of solar radiation from the surface),
r ≈ 6371 × 10
3 m is the mean Earth radius, π = 3.1415927, and σ = 5.67 × 10
−8 J·s
−1·m
−2·K
−4 is the Stefan–Boltzmann constant. Based on Equation (1), if we had no atmosphere, the Earth’s mean surface temperature would be approximately
Te = 255 K = −18 °C.
Adding gases to the atmosphere has a minimal impact on the surface solar radiation balance, but greenhouse gases in the atmosphere are known to intercept upward and downward fluxes of LWR, which raises the atmospheric temperature. Assuming a single layer atmosphere that is opaque to LWR, it can be shown that Equation (2) provides an estimate of the maximum Earth surface temperature (
Ts) assuming that all of the LWR is absorbed by the atmosphere.
Thus, if the GHG in the atmosphere was 100% efficient at intercepting LWR, the maximum surface temperature would be Ts = 303 K = 30 °C. Therefore, with current solar radiation, short waveband radiation reflection, and GHGs, the actual surface temperature of Earth falls somewhere between −18 °C and 30 °C.
Assuming a single layer leaky atmosphere, which allows for some loss of LWR back to space, it can be shown that Equation (3) provides an estimate of the Earth surface temperature (
Ts) as a function of
Te and the fraction of upward LWR absorbed by the atmosphere (
ϵ). Until the GHG concentration began to increase, about 78% of the LWR was absorbed by GHGs in the atmosphere. Therefore, using Equation (3) and
ϵ = 0.78 shows that GHG increased the Earth’s surface temperature from −18 °C to 15 °C (288 K).
Equation (3) illustrates that
Ts is affected only by
Te and
ϵ, and Equation (1) shows that
Te is affected only by the amount of incoming and reflected solar radiation. Thus, only the incoming and reflected short waveband radiation and GHG concentration determine the Earth’s global surface temperature. Since 1958, changes in
RSC and
αp were insignificant; however, the anthropogenic releases of CO
2 and other GHGs led to a rise of more than 80 ppm in atmospheric CO
2 concentration. The rising CO
2 concentration increased
ϵ and it is the most likely cause for most of the corresponding 1 °C increase in global temperature [
6]. Based on [
1], considerably more global warming is projected as CO
2 levels continue to rise. While global temperature is rising, other climate factors, e.g., humidity, precipitation, wind speed, cloudiness, and precipitation, are also changing, and the assessment of climate change impact on weather and horticultural crops should consider all factors.
4. Estimating Crop Evapotranspiration
For many decades, scientists, engineers, and irrigation managers have used ET and the water balance method to determine agricultural and urban water demand for water resource planning and delivery and for on-farm and urban irrigation management. Because ET is affected by soil, plant, and atmospheric factors, spatial ET variation is common on different scales. In some climates, one can estimate ET using weather data of large areas, e.g., 50–100 km, but in other locations, microclimates can limit weather-based estimates of ET to small areas, e.g., less than 5 km. In general, the most common practice is to estimate “potential” or “energy-limited” crop evapotranspiration (ETc) as the product of reference evapotranspiration (ETref) and a crop coefficient (Kc). Reference evapotranspiration is the energy-limited ET rate from a broad expanse of a well-watered vegetated surface, e.g., grass or alfalfa.
There are several methods to determine
ETref, but the most common is to monitor weather over a large grass surface and use an equation to estimate the water use for a selected reference surface. Recently, the authors of [
7,
8] recommended fixed coefficients to estimate the canopy and aerodynamic resistances as inputs to the Penman–Monteith equation [
9] to estimate
ETref for 0.12-meter-tall and 0.50-meter-tall vegetated surfaces and assigned the symbol
ETo for the short canopy and
ETr for the tall canopy. The equations were called “standardized reference evapotranspiration” because the procedures to compute the
ETo and
ETr were standardized. Strictly speaking, the
ETo and
ETr equations provide estimates of ET from a virtual surface with input coefficients to estimate the appropriate canopy and aerodynamic resistances. However, in practice, the
ETo and
ETr approximate the ET of 0.12-meter-tall cool-season grass and 0.50-meter-tall alfalfa, respectively. Once the
ETref is known, the
ETc is estimated as either
ETc =
ETo × Kc or
ETc =
ETr × Kcr, where
Kcr is specific to
ETr. Because
ETo is more widely used than
ETr, the remaining discussion will only use the
ETo method. Crop coefficients are generally determined as
Kc =
ETc/
ETo by calculating the
Kc ratio using measured daily
ETc and
ETo from weather data. Global climate change could affect either the
ETo or the
Kc, so it is important to assess the impact of climate change on both factors.
5. Reference Evapotranspiration
To assess the possible impact of climate change on
ETo, Snyder et al. [
10] used the standardized reference evapotranspiration (
ETo) equation for short canopies [
7,
8] that uses daily radiation, temperature, humidity, wind speed, and canopy resistance data to calculate
ETo. The following version of the
ETo equation [
11] was used for the analysis:
In Equation (4), ∆ (kPa·K
−1) is the slope of the saturation vapor pressure curve at the mean daily air temperature,
Rn (MJ·m
−2·d
−1) is the net radiation over well-watered grass,
G (MJ·m
−2·d
−1) is the soil heat flux density,
γ (kPa·K
−1) is the psychrometric constant,
T (°C) is the mean daily air temperature at a 1.5–2.0 m height,
u2 (m·s
−1) is the wind speed measured at 2 m above the ground, and
es and
ea (kPa) are the saturation and actual vapor pressures of the air measured at a 1.5 m height. Information on how to compute
Rn,
G, ∆, etc. is provided in [
7].
The aerodynamic resistance to sensible and latent heat transfer (
ra) occurs indirectly in two locations in Equation (4). The 0.34 in the denominator comes from the following:
In Equation (4), the right-hand side of the numerator could be written as
Therefore, the 208 coefficient is also included in the numerator of Equation (4) (within the 900). The
rc = 70 s·m
−1 was estimated from the typical stomatal resistance
rs = 100 s·m
−1 (corresponding to stomatal conductance
gs = 0.010 m·s
−1 = 10 mm·s
−1) for the actively transpiring C
3 grass leaf surface, which was estimated as half of the LAI = 2.88. Therefore, the canopy resistance for 0.12-meter-tall C
3 species grass
rc was calculated as
Long et al. [
12] studied the effect of CO
2 concentration on stomatal conductance of C
3 species plants, and the relationships they reported were used to estimate the impact of increasing CO
2 concentration on canopy resistance of the standardized reference surface. Assuming that the
rc ≈ 70 s·m
−1 applies to a 2004 CO
2 concentration of about 372 ppm, estimating a new
rc value for higher CO
2 concentration provides a method to estimate possible impacts of higher CO
2 on
ETo.
Based on eight climate models, the mean air temperature projection for California is about 2.2 °C by 2050 and 4.0 °C by 2100 [
13]. The global CO
2 concentration is projected to reach about 550 ppm by 2050 and more than 700 ppm by 2100 [
14]. Long et al. [
12] reported that stomatal conductance for many C
3 species plants decreased by about 20% when the CO
2 concentration was increased from 372 to about 550 ppm from about 200 independent measurements. Assuming this is true for the stomatal conductance of 0.12-meter-tall C
3 species grass with a stomatal resistance of 100 s·m
−1, the stomatal conductance for C
3 grass should decrease from about 10 mm·s
−1 to 8 mm·s
−1, which corresponds to
rs = 125 s·m
−1. Using the same approach used to calculate
rc in the
ETo equation [
7], the
rc for 550 ppm is calculated as
Thus, increasing CO2 concentration from 372 to 550 ppm should increase canopy resistance of 0.12-meter-tall C3 grass from 70 to 87 s·m−1.
Roderick and Farquhar [
15] observed that the global mean daily maximum temperature (
Tx) increased by approximately 0.1 °C per decade and that the daily minimum temperature (
Tn) increased by about 0.2 °C per decade, but there was no change in the vapor pressure deficit (
VPD =
es −
ea) in the few preceding decades. Cayan et al. [
13] projected a 3 °C mean temperature increase by 2050 as a worst-case scenario for California. Therefore, Snyder et al. [
10] evaluated the possible impact of climate change on
ETo by assuming that the mean air temperature would increase by 3.0 °C, with
Tn and
Tx increasing by 4.0 °C and 2.0 °C, respectively. Globally, the temperatures were rising, with no increase in relative humidity, which implies that the vapor pressure was also rising. In much of California, the mean daily dew point temperature (
Td) is often nearly equal to
Tn, so it was assumed that the
Td like
Tn would increase by 4 °C in 2050. It was also assumed that the CO
2 concentration would increase from 372 to 550 ppm by volume by 2050, which corresponds to increasing the
rc from 70 to 87 s·m
−1. It was assumed that aerodynamic resistance, which is dependent on the atmospheric stability, wind speed and plant canopy roughness, would not change.
Using 18 California weather stations from a wide range of climates, daily data from 2003 were used to calculate how the annual
ETo might change with the aforementioned climate projections. The results (
Figure 1) indicate that the climate change scenario would have little impact on annual
ETo. The annual
ETo increased slightly where there were mean wind speeds less than 1.7 m·s
−1, and it decreased for wind speeds greater than 1.7 m·s
−1. However, the absolute magnitude of variation from the current annual
ETo was small for all weather stations. Based on the regression lines in
Figure 1, the 4 °C rise in
Td and the 17 s·m
−1 increase in
rc counteracted the impact of the 3 °C mean temperature increase.
6. Crop Coefficients
Allen et al. [
7,
8,
11] reported a daily time step, standardized reference evapotranspiration equation for short canopies (Equation (7)). While the generated
ETo is actually for a virtual surface, i.e., having characteristics that determine the coefficients in Equation (4), the
ETo approximates the
ETc for a broad expanse of well-watered, cool-season grass. Allen et al. [
7,
8] also reported a daily time step equation for tall canopies (
ETr) that is expressed as
Note that Allen et al. [
7] used
rc = 45 s·m
−1 and
ra = 118/u
2 to obtain the 0.38 = 45/118 in Equation (9). The canopy resistance for about 372 ppm CO
2 was derived using Equation (10) with
rs = 66.7 s·m
−1 and LAI = 2.96.
Assuming a 20% reduction in alfalfa stomatal conductance when changing from 372 to 550 ppm CO
2 [
12], the stomatal conductance changes from 0.015 m·s
−1 to 0.012 m·s
−1, the stomatal resistance changes from 66.7 to 83.3 s·m
−1, and the canopy conductance at 550 ppm is estimated as
Strictly speaking, the
ETr equation estimates the
ETc for a virtual surface with characteristics represented by coefficients used in Equation (9), but the
ETr rates are approximately equal to the
ETc of a broad expanse of well-watered, 0.5-meter-tall alfalfa. Derivations and explanations of the
ETo and
ETr equations are addressed in [
7,
8]. To accurately estimate
ETo and
ETr, the input weather data for Equations (4)–(9) are collected over a broad expanse of well-watered grass. Derivation of the net radiation (
Rn) and ground heat flux (
G) in MJ·m
−2·d
−1 are presented in [
7]. The
T (°C) is the mean air temperature measured at a height between 1.5 and 2.0 m,
u2 (m·s
−1) is the mean daily wind speed monitored at a 2 m height, the saturation vapor pressure (
es) in kPa is calculated from
T, and the actual vapor pressure (
ea) in kPa is measured at the same height as
T. The slope of the saturation vapor pressure (∆) in kPa·°C
−1 and the psychrometric constant
γ in kPa·°C
−1 are computed as described in [
7].
While little change in
ETo rates is expected due to global climate change, crop coefficient (
Kc) values might be affected depending on how climate factors change in the future. To evaluate the effect of changing weather variables on
Kc values, an analysis was done using data from 49 California Irrigation Management Information System (CIMIS) weather stations [
16]. Equations (4) and (9) were used to compute
ETo and
ETr using canopy resistances corresponding to both 372 and 550 ppm CO
2. The stomatal conductances
gs = 0.010 m·s
−1 (grass) and
gs = 0.015 m·s
−1 (alfalfa) were reduced by 20% to 0.008 m·s
−1 (grass) and 0.012 m·s
−1 (alfalfa) at 550 ppm CO
2 based on [
12]. The
Kc for alfalfa was calculated for 372 ppm CO
2 for all 49 stations using the mean climate data from July 2003. For 550 ppm CO
2, the
Kc for alfalfa was calculated using the mean climate data from July 2003, but with the monthly mean daily maximum, minimum, and dew point temperatures increased by 2 °C, 4 °C, and 4 °C, respectively. The wind speed and equations for aerodynamic resistance were not changed.
A plot of the 550 ppm versus 372 ppm
Kc values (
Figure 2) indicates that the crop coefficients are likely to decrease slightly for a wide range of climates. The biggest
Kc decrease was about 0.03 when
Kc ≈ 1.10, and the smallest decrease was about 0.01 for
Kc ≈ 1.40. The differences were approximately 0.01 when
Kc values were high. The response of
Kc to the projected climate change is most likely related to the alfalfa canopy being more coupled with the environment than the grass canopy, which has a higher aerodynamic resistance. Plant canopies that are more coupled to the environment are more likely to exhibit a reduction in transpiration rate than a canopy that is more controlled by the aerodynamic resistance, e.g., grass. This analysis provides some evidence that coupling with the environment might lead to reductions in crop coefficients due to global climate change.
While changes in the CO2 concentration can have an effect on the canopy resistance, this analysis showed that the same percentage decrease stomatal conductance can lead to a bigger reduction in transpiration if the aerodynamic resistance is lower and the canopy is more coupled with the environment. There is a lack of information on how crop coefficients of orchard and vine crops might respond to climate change, but the alfalfa example in this paper provides some insight. Since taller rougher canopies, e.g., orchards and vineyards, have considerably lower aerodynamic resistance than alfalfa, it is likely that the Kc values for orchard and vine crops might decrease even more.
For both C
4 and C
3 species, stomatal conductance is reduced by the increasing CO
2 concentration external to leaves [
17]. However, small differences in C
4 and C
3 stomatal conductance responses to CO
2 concentration have been reported for grasses with the conductance differences decreasing at higher CO
2 concentrations [
18]. On the other hand, grasses are more decoupled from the environment than taller, rougher plant canopies, so the
Kc response of grass species due to a projected climate is likely to be smaller than the
Kc response of taller rougher plant canopies with lower aerodynamic resistance.