Net irrigation water requirements for major irrigated crops with variation in evaporative demand and precipitation in southern Alberta

Abstract

Weather variability has a profound influence on crop and irrigation water requirements. Estimates of crop evapotranspiration (ET_c) and net irrigation water requirements are needed for water allocation, risk management and irrigation system planning. Seasonal ET_c and net irrigation water requirement estimates based on the standardized Penman-Monteith method were examined through frequency analysis of historical weather data. Seasonal ET_c calculated using the Penman-Monteith equation is based on daily solar radiation, air temperature, relative humidity, and wind speed. Historical weather data from 1983 to 2012 at Lethbridge and Vauxhall were used to determine seasonal ET_c, seasonal precipitation and net irrigation water requirements for 11 major (most prevalent) irrigated crops in southern Alberta. Seasonal ET_c was consistently greater at Lethbridge than Vauxhall, whereas seasonal precipitation was generally less at Vauxhall than Lethbridge for all major crops. Mean seasonal ET_c ranged from 355 mm for barley silage at Vauxhall to 728 mm for alfalfa hay at Lethbridge at a 10% chance of exceedance. Mean net irrigation water requirements ranged from 273 mm for barley silage at Vauxhall to 526 mm for alfalfa hay at Lethbridge at a 10% chance of exceedance. Area-weighted seasonal ET_c demand within the irrigation districts is currently about 500 mm (2.8 billion m³) and the net irrigation water requirement within the irrigation districts is at least 380 mm (2.1 billion m³) at a 10% chance of exceedance. Annual gross diversion requirements for the irrigation districts could approach the licensed water allocation limit of 3.45 billion m³ at a 10% chance of exceedance when conveyance losses, irrigation system application efficiencies, and current irrigation management practices are considered. The frequency with which annual gross irrigation water requirements approach or exceed this licensed water allocation limit may increase in the future with climate change in southern Alberta.

Abstract

La variabilité des conditions météorologiques influence grandement les besoins en eau des cultures et pour l’irrigation. Les estimations d’évapotranspiration des cultures et les besoins net en eau d’irrigation sont nécessaires pour l’attribution de l’eau, gestion des risques et la planification stratégique de développement de l’irrigation. Les estimations d’évapotranspiration saisonnière et les besoins net en eau d’irrigation basés sur la méthode normalisée de Penman-Monteith ont été comparés aux données météorologiques historiques. Les estimations d’évapotranspiration saisonnière calculées à l’aide de l’équation de Penman-Monteith reposent sur le rayonnement solaire, la température de l’air, l’humidité relative et la vitesse du vent. Les données météorologiques historiques de 1983 à 2012 pour Lethbridge et Vauxhall ont servi à déterminer l’évapotranspiration saisonnière, les précipitations saisonnières et les besoins net en eau d’irrigation pour les 11 cultures irriguées les plus répandues dans le sud de l’Alberta. L’évapotranspiration saisonnière moyenne était constamment supérieure à Lethbridge comparativement à Vauxhall, tandis que les précipitations saisonnières étaient généralement inférieures à Vauxhall qu’à Lethbridge pour les principales cultures. L’évapotranspiration saisonnière moyenne variait de 355 mm pour l’orge fourragé à Vauxhall à 728 mm pour la luzerne fourragère à Lethbridge pour une chance de dépassement de 10%. Les besoins net en eau d’irrigation variaient de 273 mm pour l’orge fourragé à Vauxhall à 526 mm pour la luzerne fourragère à Lethbridge pour une chance de dépassement de 10%. L’évapotranspiration saisonnière dans les districts d’irrigation correspond actuellement à environ 500 mm (2,8 milliards de m³) d’eau et le besoin net en eau d’irrigation dans les districts d’irrigation est d’au moins 380 mm (2,1 milliards de m³) pour une chance de dépassement de 10%. Le volume de diversion brute annuelle nécessaire pour combler les besoins des districts d’irrigation pourrait s’approcher de la limite de la licence d’allocation en eau de 3,54 milliards de m³ pour une chance de dépassement de 10% lorsque les pertes lors de l’adduction, l’efficacité des systèmes d’irrigation et les pratiques actuelles de gestion de l’irrigation sont considérées. La fréquence à laquelle les besoins annuels bruts en eau d’irrigation atteignent ou dépassent la limite d’allocation en eau sous licence peut augmenter dans l’avenir en considérant les changements climatiques potentiels pour le sud d’Alberta.

Introduction

Variability in weather parameters has a profound influence on crop and irrigation water requirements. Crop water requirements per unit area are defined as the depth of water needed to meet the water used through evapotranspiration (Pereira and Alves 2005). The net irrigation water requirement is the depth of water needed to fulfill the crop water requirement in excess of any effective precipitation for a disease-free crop growing in large fields under nonrestricting soil and soil water conditions and under adequate fertility (Allen et al. 2011). Gross irrigation requirements account for the application efficiencies of the irrigation systems used and may include additional applications for leaching of salts or to offset nonuniformity of water application (Allen et al. 2011).

The current licensed water allocation for the irrigation districts in Alberta is 3.45 billion m³ and the median annual natural flow volume from the South Saskatchewan River Basin is about 8.8 billion m³ (AMEC 2009). Natural flow (discharge) is flow that is not noticeably affected by direct human activities, including reservoir operation, water withdrawals, diversions or releases (AMEC 2009). Alberta is authorized to consume or store half of the apportionable flow of the South Saskatchewan and Red Deer Rivers under the 1969 Master Agreement on Apportionment (AMEC 2009). Water storage infrastructure in the South Saskatchewan River Basin includes at least 50 storage reservoirs with a live storage capacity of about 3 billion m³ (Alberta Agriculture and Rural Development 2012).

Weather-related variables that determine irrigation requirements include crop evapotranspiration (ET_c) demand and the amount and timing of precipitation (Andales 2009). The ET_c demand varies daily according to crop growth stages, amount and frequency of wetting of the soil surface, environmental conditions and crop management (Allen et al. 2011). A Penman-Monteith evapotranspiration (ET) equation has been standardized to estimate reference ET of 12-cm-tall cool-season grass and 50-cm-tall alfalfa (Allen et al. 2005) and is currently the most common method used to estimate reference ET (Allen et al. 2011). Alfalfa is used for estimation of reference ET in southern Alberta, and locally-derived crop coefficients based on cumulative growing-degree days have been developed for determination of ET_c (Bennett and Harms 2011).

Effective precipitation is the portion of precipitation that infiltrates into the soil, is retained in the root zone and is used to meet crop water requirements (Allen et al. 2011). Surface runoff and deep percolation losses during precipitation events are affected by precipitation intensity and duration, antecedent soil water content, land cover, land slope, soil texture, soil structure, sealing and crusting of the soil surface, and tillage practices (Allen et al. 2007). Thus, quantification of the amount of effective precipitation from historical weather data has a high degree of uncertainty.

Earlier crop and irrigation water requirement studies in Alberta include work by Steed and Ulrickson (1971), using the Blaney-Criddle method (Blaney and Criddle 1962), and by Underwood McLellan Ltd. (1982), based on pan evaporation for reference ET and field experiments with selected crops by Agriculture Canada from 1949 to 1968 (e.g. Hobbs and Krogman 1968). More recent water resources planning in southern Alberta has been based on determination of ET_c and irrigation requirements based on a modified Priestley-Taylor empirical method, which is based on maximum and minimum air temperature and solar radiation (Irrigation Water Management Study Committee 2002). Since the ET process is physically limited by available energy and is controlled by energy exchange at the vegetation surface, the Penman-Monteith method has been widely adopted for determination of ET_ref (Allen et al. 1998, 2005, 2011). The Food and Agricultural Organization of the United Nations and the American Society of Civil Engineers have recommended that weather parameters required by the physically-based Penman-Monteith method be estimated from available weather data rather than using an empirical method to determine ET_ref (Allen et al. 2011). Standardized procedures have been developed to estimate these missing weather parameters (Allen et al. 2005).

Historical weather data can be used to estimate the chance of occurrence that specified ET demand, precipitation and deficit amounts will occur in a given area. Frequency analysis of historical weather data can be performed using a series of ranked values plotted against the chances that these values will be exceeded (Chow et al. 1988; Andales 2009). Different levels of risk can be evaluated by selecting a chance of occurrence and then determining the corresponding ET_c or deficit that will likely occur. This approach assumes equally-spaced data, which rarely occurs with hydrology data.

Risk analysis may also be performed using Monte Carlo simulations (Aggarwal 1995; Rubinstein and Kroese 2008; Maneta et al. 2009; Tao et al. 2009; Maeda et al. 2011; Graveline et al. 2012; Spank et al. 2013) based on historical weather data. In a Monte Carlo simulation, values are sampled randomly hundreds or thousands of times from input probability distributions, each time using different randomly-selected values, resulting in a probability distribution of possible outcome values. Results indicate what could happen and the likelihood that it will occur.

The objective of our study was to examine crop and irrigation water requirements for major irrigated crops in southern Alberta through risk analysis of historical weather data. Physically-based estimates of crop evapotranspiration (ET_c) and net irrigation water requirements determined using the standardized Penman-Monteith method are needed for water allocation, risk management and irrigation system planning.

Materials and methods

Weather data

Solar radiation, air temperature, relative humidity and wind speed data from Agriculture and Agri-food Canada weather stations at Lethbridge and Vauxhall, Alberta (Figure 1), were used to calculate daily Penman-Monteith reference ET (alfalfa) and ET_c for 11 major (most prevalent) irrigated crops in southern Alberta from 1983 to 2012 (Equation 1): (1)

Figure 1. Weather stations at Lethbridge and Vauxhall in southern Alberta.

where ET_c = crop evapotranspiration (mm), K_c = daily crop coefficient and ET_ref = reference evapotranspiration (mm).

Meteorological sensors, measurement techniques and site characteristics were consistent during this time interval, except for automation of instrumentation at Lethbridge in 1986 and Vauxhall in 1987. Seasonal ET_c was calculated using daily crop coefficients (Bennett and Harms 2011) and by accumulating daily ET_c values for the growing season each year. Daily crop coefficients change during the growing season as plants grow and develop, as the fraction of ground cover changes, as the wetness of the soil surface changes and as plants age and mature (Allen et al. 2007). The seasonal ET_c for alfalfa and grass hay was determined from 1 May to 30 September, and seasonal ET_c for annual crops was estimated from 15 May until maturation predicted by cumulative growing-degree days or when a killing frost occurred. Crop water use outside these intervals was generally less than 5% of the seasonal amount. Wever et al. (2002) found that ET_c for grassland near Lethbridge during a 3-year interval (1998–2000) was less than 0.5 mm per day in the winter months.

Historical meteorological data sets needed to determine Penman-Monteith reference ET and ET_c were not available from other weather stations within the irrigated areas of southern Alberta, except for the last decade, so data from these other weather stations were not analyzed due to the short period of record. Historical precipitation data at Lethbridge and Vauxhall were not adjusted to “effective” precipitation amounts due to the high degree of uncertainty and lack of data needed for this correction. Data for soil moisture content in the spring prior to planting and in the fall after harvest were also not available. Net irrigation water requirements for each crop at these locations were estimated using Equation (2): (2)

where IR_net = net irrigation requirement (mm), ET_c = crop evapotranspiration (mm) and P_s = seasonal precipitation (mm).

Major irrigated crops

Cereal grains (barley, hard and soft spring wheat) and canola were grown on about 242,930 ha within the irrigation districts in southern Alberta in 2011 (Table 1). Forage crops (alfalfa hay, barley silage, corn silage and grass hay) were produced on another 132,440 ha and specialty crops (dry bean, potato and sugar beet) were grown on about 44,230 ha within the irrigation districts in 2011 (Alberta Agriculture and Rural Development 2012). These major irrigated crops represented 76% of the assessed area within the irrigation districts in 2011.

Table 1. Area of major crops within the irrigation districts of southern Alberta in 2011.

Crop	Latin name(s)	Irrigated area, ha^*
Alfalfa (hay)	Medicago sativa	59,710
Barley	Hordeum vulgare	51,170
Barley silage	Hordeum vulgare	29,750
Canola	Brassica napus	82,230
Corn silage	Zea mays	22,190
Dry bean	Phaseolus vulgaris	12,290
Grass (hay)	Bromus inermis; Dactylis glomerata; Phleum pratense	20,790
Hard spring wheat	Triticum aestivum; Triticum turgidum	104,460
Potato	Solanum tuberosum	17,490
Soft spring wheat	Triticum aestivum	5,070
Sugar beet	Beta vulgaris	14,450
Subtotal		419,600
Total assessed area		554,600

^* Source: Alberta Agriculture and Rural Development (2012).

Frequency analysis

Seasonal ET_c, seasonal precipitation and net irrigation requirements (deficits) for each major crop at Lethbridge and Vauxhall (Figure 1) were ranked from greatest to least values versus the chances that these values would be exceeded. The Weibull formula (Equation 3; Chow et al. 1988) was used to estimate a given value’s chance of exceedance (Pe): (3)

where Pe = chance of exceedance, m = rank of a value in a list arranged from greatest to least and n = total number of observations or values. The greatest value was given the rank of 1 (m = 1) and the least value was assigned the rank of n (m = n).

Risk analysis

Seasonal ET_c, seasonal precipitation and net irrigation requirement data for each of the 11 major irrigated crops at Lethbridge and Vauxhall were compiled at chance of exceedance values of 10, 50 and 90% for risk analysis. A given parameter value at a 90% chance of exceedance will likely be exceeded 90% of the time, whereas a parameter value at a 10% chance of exceedance will be exceeded about 10% of the time. Previous water management planning studies in southern Alberta have used the 90^th percentile irrigation demand for determination of licensed allocations (Irrigation Water Management Study Committee 2002).

Preliminary analyses of the data sorted from greatest to least values indicated that the data had a skewed distribution with some relatively large and small values. Fitting the data to a probability distribution to estimate the chance of exceedance was thus not practical with only 30 years of data from each site. Monte Carlo simulations (Rubinstein and Kroese 2008) were subsequently performed to obtain estimates of ET_c, precipitation and net irrigation water requirements for 11 crops at 10, 50 and 90% chances of exceedance using the 30 years of data observed at Lethbridge and Vauxhall. For each site, crop and selected chance of exceedance, a data set was generated by randomly removing five data points from the observed data set, and then the UNIVARIATE procedure (SAS Institute Inc. 2012) was used to calculate ET_c, precipitation and net irrigation water requirement values for the selected chance of exceedance. Five observations were randomly removed using a standard random sampling technique. Specifically, each observed data value was paired with a generated random number. The data were then sorted by the random number, which caused the observed data to be randomly shuffled, and then the first five observations were discarded. This process was repeated 500 times and the UNIVARIATE procedure was used to calculate the mean, standard error, minimum and maximum, and 95% confidence intervals for each parameter at the selected chance of exceedance. The 500 iterations provide information on variation about the mean estimate that one would expect in real-world situations. Auto-correlations among the weather data were evaluated by correlating data from a given year to data from one (lag1) and two (lag2) prior years using the CORR procedure (SAS Institute Inc. 2012).

Gross irrigation district water requirements

The mean seasonal ET_c and net irrigation water requirements for Lethbridge and Vauxhall were area-weighted by crop type and extrapolated to the total area assessed for irrigation within the irrigation districts in 2011 (554,613 ha) to obtain a gross estimate of overall seasonal ET_c and net irrigation water requirements for all of the irrigation districts in Alberta, expressed on a volume basis, at chance of exceedance values of 10, 50 and 90% (Equation 4). (4)

where V = volume (m³), ET_c = crop evapotranspiration (mm), IR_net = net irrigation requirement (mm) and A = area (ha).

Lethbridge and Vauxhall are in the central portion of the irrigated area in southern Alberta (Figure 1). The irrigated area to the east is generally warmer and drier, whereas the irrigated areas to the north and west are cooler and wetter. Gross irrigation water requirements for the irrigation districts were estimated using an overall area-weighted irrigation system application efficiency of 77%. Area weighting was based on current estimates of system application efficiencies (Alberta Agriculture and Rural Development 2011) and the areas irrigated by different irrigation methods in 2011 (Alberta Agriculture and Rural Development 2012).

Results and discussion

Seasonal crop evapotranspiration

Mean seasonal ET_c values were the least for barley silage and the greatest for alfalfa hay at Lethbridge and Vauxhall at chances of exceedance of 10, 50 and 90% (Table 2). Monte Carlo simulations with 500 iterations resulted in small standard error values and narrow 95% confidence intervals, so only the minimum and maximum values for each chance of exceedance level were reported. The mean seasonal ET_c value for barley silage of 297 mm would be exceeded about 90% of the time and a value of 341 mm would be exceeded about 50% of the time, whereas a value of 399 mm would be exceeded about 10% of the time at Lethbridge. Mean seasonal ET_c for barley silage at Vauxhall ranged from 273 mm at a 90% chance of exceedance to 355 mm at a 10% chance of exceedance. In contrast, mean seasonal ET_c for alfalfa ranged from 506 mm at a 90% chance of exceedance to 728 mm at a 10% chance of exceedance at Lethbridge. Mean seasonal ET_c for alfalfa at Vauxhall was 498 mm at a 90% chance of exceedance and 672 mm at a 10% chance of exceedance. Lag1 correlations for seasonal ET_c were significant for all crops except barley silage and corn silage at Lethbridge, and lag2 correlations were significant for alfalfa at Lethbridge, but these auto-correlations were all low (Table 3). For example, the greatest auto-correlation for alfalfa at Lethbridge was 0.47, which means that 22% of the variation in ET_c in one year can be accounted for by the ET_c values in the previous year and 78% was unaccounted for. Auto-correlations for seasonal ET_c were not significant for all crops at Vauxhall. These low or insignificant correlations indicate that the ET_c value observed in any given year is not useful for predicting ET_c values observed the following two years and, for all practical purposes, these auto-correlations are trivial.

Table 2. Mean (range) crop evapotranspiration (mm) from 1983 to 2012 at different chances of exceedance for major irrigated crops at Lethbridge and Vauxhall.

Crop	Lethbridge			Vauxhall
	Chance of exceedance, %
	10	50	90	10	50	90
Alfalfa (hay)	728(706–743)	616(596–638)	506(487–566)	672(652–680)	586(571–603)	498(492–521)
Barley	437(398–439)	365(364–367)	329(328–352)	380(368–381)	337(334–351)	300(298–309)
Barley silage	399(355–401)	341(333–348)	297(288–318)	355(346–357)	308(301–316)	273(270–282)
Canola	467(434–473)	389(387–392)	359(354–373)	404(388–405)	362(358–378)	325(321–338)
Corn silage	537(506–541)	464(457–468)	385(363–429)	472(453–475)	427(419–433)	381(379–405)
Dry bean	475(433–478)	407(404–409)	356(352–383)	413(406–414)	375(366–378)	338(335–345)
Grass (hay)	529(490–536)	454(442–461)	390(375–424)	501(481–506)	444(440–454)	387(384–415)
Hard wheat	538(500–545)	453(450–456)	419(416–438)	468(453–469)	420(414–435)	377(372–390)
Potato	581 (551–598)	490 (476–497)	386 (356–435)	530(510–535)	466(453–470)	387(384–417)
Soft wheat	539(510–545)	453(450–456)	418(416–438)	468(458–469)	420(414–435)	377(372–390)
Sugar beet	644(615–648)	553(545–566)	447(413–498)	589(561–594)	522(513–527)	443(441–478)

Table 3. Auto-correlations for crop evapotranspiration (ET_c), growing-season precipitation, and net irrigation water requirements (mm) from 1983 to 2012 for major irrigated crops at Lethbridge and Vauxhall.

	Lethbridge						Vauxhall
Crop	ETc		Precipitation		Net irrigation		ETc		Precipitation		Net irrigation
	lag1	lag2	lag1	lag2	lag1	lag2	lag1	lag2	lag1	lag2	lag1	lag2
Alfalfa (hay)	0.47^*	0.42^*	–0.23	–0.18	0.07	0.13	0.26	0.12	–0.24	–0.13	0.03	–0.03
Barley	0.41^*	0.23	0.03	–0.13	0.20	0.00	0.09	–0.06	0.02	–0.21	0.10	–0.18
Barley silage	0.36	0.27	0.09	–0.07	0.22	0.05	0.09	–0.12	0.08	–0.15	0.15	–0.11
Canola	0.41^*	0.16	–0.01	–0.12	0.17	0.02	0.12	–0.04	0.00	–0.21	0.12	–0.16
Corn silage	0.36	0.22	–0.22	–0.24	0.00	0.03	0.32	0.11	–0.24	–0.20	–0.03	–0.10
Dry bean	0.42^*	0.19	–0.22	–0.24	0.00	0.03	0.26	0.02	–0.26	–0.22	–0.06	–0.11
Grass (hay)	0.38^*	0.33	–0.23	–0.18	–0.06	0.03	0.23	0.17	–0.24	–0.13	–0.05	–0.07
Hard wheat	0.39^*	0.13	–0.15	–0.19	0.09	0.02	0.16	–0.04	0.00	–0.21	0.12	–0.13
Potato	0.37^*	0.34	–0.22	–0.25	0.02	0.08	0.17	0.10	–0.22	–0.20	–0.04	–0.08
Soft wheat	0.39^*	0.13	–0.07	–0.20	0.15	0.01	0.16	–0.04	0.00	–0.18	0.11	–0.12
Sugar beet	0.39^*	0.34	–0.22	–0.25	0.05	0.06	0.20	0.13	–0.22	–0.19	–0.01	–0.08

^* Significant at P < 0.05.

Variation in evaporative demand was greater for perennial forage and long growing season annual crops (corn silage, potato, sugar beet) than the other crops at Lethbridge (Table 2). Similar differences in variability in ET_c were evident among crops at Vauxhall, except for corn silage, which was comparable to spring wheat. The greater mean seasonal ET_c values at Lethbridge than Vauxhall likely resulted from greater evaporative demand due to differences in wind-related demand (Figure 2).

Figure 2. Seasonal wind measured at Lethbridge and Vauxhall from 1983 to 2012.

Mean seasonal ET_c at a 50% chance of exceedance in this study was considerably less than growing season water use previously reported (Table 4) for all major crops except dry bean and hard spring wheat, which had comparable seasonal ET_c. The seasonal ET_c reported by Alberta Agriculture and Rural Development (2008) for barley (460 mm), barley silage (430 mm), canola (480 mm) and grass hay (560 to 610 mm) exceeded our mean seasonal ET_c at a 10% chance of exceedance at Lethbridge and Vauxhall.

Table 4. Crop evapotranspiration (mm) for major irrigated crops in southern Alberta.

Crop	Sonmor (1963)	Underwood McLellan(1982)	Alberta Agriculture and Rural Development(2008)
Alfalfa (hay)	648	660	680
Barley	409	410	460
Barley silage	-	-	430
Canola	-	410	480
Corn silage	373	530	510
Dry bean	-	-	380
Grass (hay)	599	562	560–610
Hard spring wheat	462	510	460
Potato	505	580	560
Soft spring wheat	493	-	480
Sugar beet	546	600	560

Seasonal precipitation

Seasonal precipitation for barley silage ranged from 44 mm in 1985 to 369 mm in 1993 at Lethbridge and from 39 mm in 1985 to 251 mm in 2005 at Vauxhall. Minimum seasonal precipitation for various crops occurred in 1985, 2000 and 2001 at Lethbridge and in 1985, 1996 and 2001 at Vauxhall. Maximum seasonal precipitation was observed in 1993 and 2005 at Lethbridge and in 2002 and 2005 at Vauxhall. Mean seasonal precipitation for silage barley of 73 mm would be exceeded about 90% of the time and a seasonal precipitation value of 299 mm would be exceeded about 10% of the time at Lethbridge (Table 5). The mean seasonal precipitation for silage barley at Vauxhall ranged from 61 mm at a 90% chance of exceedance to 225 mm at a 10% chance of exceedance. Mean seasonal precipitation for alfalfa or grass ranged from 152 mm at Lethbridge and 149 mm at Vauxhall at a 90% chance of exceedance to 409 mm at Lethbridge and 320 mm at Vauxhall at a 10% chance of exceedance. Lag1 and lag2 correlations for growing-season precipitation were not significant for all crops at Lethbridge and Vauxhall (Table 3).

Table 5. Mean (range) growing-season precipitation (mm) from 1983 to 2012 at different chances of exceedance for major irrigated crops at Lethbridge and Vauxhall.

Crop	Lethbridge			Vauxhall
	Chance of exceedance, %
	10	50	90	10	50	90
Alfalfa (hay)	409(313–459)	229(218–243)	152(148–169)	320(260–333)	202(185–217)	149(147–163)
Barley	332(246–360)	182(171–197)	78(75–94)	236(196–251)	152(138–164)	64(59–112)
Barley silage	299(246–321)	168(158–177)	73(72–83)	225(179–240)	140(124–154)	61(59–97)
Canola	352(265–373)	192(172–204)	79(76–93)	248(224–254)	155(145–166)	71(67–115)
Corn silage	386(283–431)	207(195–217)	130(126–163)	282(242–294)	172(166–200)	122(110–151)
Dry bean	388(278–431)	207(195–217)	126(121–153)	282(242–294)	173(166–200)	120(103–151)
Grass (hay)	411(313–459)	230(218–243)	152(148–168)	320(260–333)	202(185–217)	149(147–165)
Hard wheat	365(277–394)	201(188–206)	87(82–126)	274(206–293)	163(157–170)	75(73–120)
Potato	389(290–431)	207(195–217)	130(126–163)	283(236–294)	175(167–200)	123(110–161)
Soft wheat	364(296–394)	196(181–205)	82(80–106)	267(220–293)	159(155–169)	71(67–120)
Sugar beet	385(290–431)	207(195–217)	129(126–163)	282(242–294)	174(167–200)	123(110–151)

Net irrigation water requirements

Total seasonal precipitation was sufficient for all crops except alfalfa in 1993 and/or 2005 at Lethbridge and net irrigation water requirements were least in 2002 and/or 2005 at Vauxhall. Maximum net irrigation water requirements at Lethbridge ranged from 375 mm for barley silage in 1985 to 738 mm for alfalfa at Lethbridge in 2001 and from 329 mm for barley silage in 1996 to 619 mm for alfalfa at Vauxhall in 2001. The greatest net irrigation water requirements for crops were observed at Lethbridge in 2001, except for barley in 2000 and barley silage in 1985, and for all crops at Vauxhall in 1996 or 2001.

Mean net irrigation water requirements for barley silage ranged from 26 mm at a 90% chance of exceedance to 318 mm at a 10% chance of exceedance at Lethbridge, and from 61 mm at a 90% chance of exceedance to 273 mm at a 10% chance of exceedance at Vauxhall (Table 6). In comparison, mean net irrigation water requirements for alfalfa would be greater than 140 mm about 90% of the time and greater than 526 mm about 10% of the time at Lethbridge, and would be greater than 182 mm about 90% of the time and greater than 520 mm about 10% of the time at Vauxhall. Mean net irrigation water requirements were similar for all crops at a 50% chance of exceedance at Lethbridge and Vauxhall. The mean net irrigation water requirements for all crops except grass hay were greater at Lethbridge than Vauxhall at a 10% chance of exceedance, whereas mean net irrigation water requirements for all crops were greater at Vauxhall than Lethbridge at a 90% chance of exceedance. Lag1 and lag2 correlations for net irrigation water requirements were not significant for all crops at Lethbridge and Vauxhall (Table 3).

Table 6. Mean (range) net irrigation water requirements (mm) from 1983 to 2012 at different chances of exceedance for major irrigated crops at Lethbridge and Vauxhall.

Crop	Lethbridge				Vauxhall
	Chance of exceedance, %
	10		50	90	10	50	90
Alfalfa (hay)		526(478–543)	388(370–420)	140(122–253)	520(472–541)	393(384–400)	182(165–276)
Barley		347(289–358)	178(160–203)	31(24–107)	293(244–307)	200(170–211)	72(66–127)
Barley silage		318(259–328)	176(165–185)	26(21–74)	273(231–286)	181(162–189)	61(54–110)
Canola		361(311–373)	193(172–219)	40(36–135)	316(255–329)	218(194–224)	82(76–127)
Corn silage		366(350–369)	241(222–270)	41(28–190)	327(290–330)	245(219–262)	106(87–146)
Dry bean		309(255–313)	186(168–226)	18(0–126)	271(244–274)	196(175–207)	66(56–130)
Grass (hay)		338(302–353)	226(215–239)	17(1–117)	352(314–371)	251(235–256)	66(53–156)
Hard wheat		415(373–425)	250(224–279)	89(78–172)	374(313–393)	261(250–281)	108(102–185)
Potato		394(359–401)	279(259–309)	40(22–147)	383(348–399)	279(267–296)	110(82–180)
Soft wheat		429(373–446)	251(233–279)	89(78–178)	376(314–393)	270(251–282)	115(103–185)
Sugar beet		464(435–468)	344(314–353)	102(81–186)	428(397–432)	332(317–356)	167(142–240)

Gross irrigation water requirements

Overall area-weighted seasonal ET_c within the irrigation districts ranged from 382 mm at a 90% chance of exceedance to 502 mm at a 10% chance of exceedance (Table 7). Crop water requirements would be 2.12 billion m³ or greater about 90% of the time, 2.41 billion m³ or more about half the time and 2.79 billion m³ or greater about 10% of the time.

Table 7. Crop evapotranspiration (ET_c) and net irrigation water requirements within the irrigation districts based on weather variability from 1983 to 2012.

	Chance of exceedance, %
Parameter	10		50		90
ETc (mm)	502		435		382
ETc (billion m^3)	2.79		2.41		2.12
Net irrigation requirement (mm)	379		252		85
Net irrigation requirement (billion m^3)	2.10		1.40		0.47

Net irrigation water requirements within the irrigation districts ranged from 85 mm at a 90% chance of exceedance to 379 mm at a 10% chance of exceedance (Table 7). Overall net irrigation water requirements would be 0.47 billion m³ or greater about 90% of the time, 1.40 billion m³ or more about 50% of the time and 2.1 billion m³ or greater about 10% of the time.

Weather variability is one of many factors that determine the amount of water diverted annually for irrigation in the South Saskatchewan River Basin. Other factors that influence the annual net irrigation water requirements include the amount and level of depletion of soil water storage, which would reduce net irrigation requirements, the amount of over-winter precipitation that contributes to soil water storage (Hobbs and Krogman 1971; Granger et al. 1984; Gray et al. 1985), and the effectiveness of precipitation in relation to surface runoff or deep percolation losses. The effectiveness of on-farm irrigation systems and irrigation management practices in meeting crop water requirements also changes annual irrigation requirements.

Variability in evaporative demand and precipitation has resulted in a wide range of irrigation water requirements for major irrigated crops. Area-weighted median (50% probability) net irrigation water requirements of 252 mm (Table 7) would require a mean gross irrigation water application within the irrigation districts of about 327 mm per year for optimum crop production, depending on the individual crops grown and irrigation system application efficiencies. About 10% of the time or less, area-weighted net irrigation water requirements of approximately 379 mm would necessitate gross irrigation water applications of about 492 mm for optimum production. The latter gross irrigation requirements are similar to the 90^th percentile area-weighted optimum farm irrigation demand of 500 mm reported by the Irrigation Water Management Study Committee (2002). Gross diversions include conveyance losses of 100 to 130 mm, volumes used for reservoir filling, and water for other licensed allocations that are supplied through irrigation district conveyance infrastructure (Alberta Agriculture and Rural Development 2012). Given current area-weighted irrigation system application efficiencies of 77% and current irrigation management practices that meet about 90% of crop requirements (Nitschelm et al. 2011), gross diversion requirements for the irrigation districts could approach the licensed water allocation limit of 3.45 billion m³ about 10% of the time. Effective use of stored water greatly reduces the chance of occurrence of annual gross diversion volumes of this magnitude. Some irrigated areas within the irrigation districts are not supported by stored water and these areas are at great risk of water shortages when water supplies are limited.

Mean gross irrigation diversions from 1976 to 2007 were about 66% of the total current irrigation district licensed allocation, ranging from about 41% in 1978 to 83% in 1988 (AECOM 2009). Irrigation districts have recently diverted about 1.0 to 2.8 billion m³ annually (2000 to 2011), with a mean annual gross diversion volume of 1.9 billion m³ during the same time interval (Alberta Agriculture and Rural Development 2012). From 1976 to 2006, Alberta passed an average of 81% of the apportionable flow to Saskatchewan, compared to the 50% required, with surplus deliveries ranging from 350 million m³ in 2001 to 5.5 billion m³ in 2005 and a mean surplus of almost 2.6 billion m³ (AMEC 2009).

Performance criteria previously used in Alberta for irrigation development planning included a growing season irrigation deficit of 75 mm or greater less than 20% of the time and a deficit of 150 mm or greater less than 10% of the time (Irrigation Water Management Study Committee 2002). Financial consequences of water supply deficits less than 100 mm (70 to 75 mm deficits at the crop level) were not considered serious for most producers (Irrigation Water Management Study Committee 2002).

Climate change may have a significant influence on ET_c, precipitation and net irrigation water requirements in the future (Supit et al. 2010). Cutforth et al. (1999) reported changes in climate in southwestern Saskatchewan during the past 50 years, with an increase in winter and spring temperatures, a decrease in snowfall amounts, an earlier start to spring runoff and a lower portion of annual precipitation as snow. A decrease in incoming solar radiation and wind speed was also detected during the summer months since the mid-1970s (Cutforth 2000). Trenberth (2011) observed increased precipitation amounts on land at higher latitudes in North America during the twentieth century, with an increase in heavy rains in most areas in the last three decades of the century. The frequency and magnitude of drought with climate change on the Canadian prairies (Axelson et al. 2009; Bonsal et al. 2013; Fleming and Sauchyn 2013) must also be considered in assessment of gross irrigation water requirements for major irrigated crops in southern Alberta.

Conclusions

Seasonal ET_c based on Penman-Monteith evaporative demand was generally greater at Lethbridge than Vauxhall, whereas seasonal precipitation was less at Vauxhall than Lethbridge for all major crops. Median (50% chance of exceedance) seasonal ET_c values based on Penman-Monteith evaporative demand for all major crops at Lethbridge and Vauxhall were generally less than values previously reported in southern Alberta. Mean seasonal ET_c at a 10% chance of exceedance ranged from 355 mm for barley silage at Vauxhall to 728 mm for alfalfa hay at Lethbridge. Net irrigation water requirements at a 10% chance of exceedance ranged from 273 mm for barley silage at Vauxhall to 526 mm for alfalfa hay at Lethbridge. The area-weighted seasonal ET_c demand within the irrigation districts is currently about 500 mm or 2.8 billion m³ and the net irrigation water requirement within the irrigation districts is equal to or greater than 380 mm or 2.1 billion m³ about 10% of the time. When conveyance losses, irrigation system application efficiencies and current irrigation management practices are considered without the benefits of stored water, gross diversion requirements for the irrigation districts could approach the licensed water allocation limit of 3.45 billion m³ about 10% of the time. The frequency and magnitude of drought with climate change in southern Alberta may increase the frequency that this licensed water allocation limit is approached or exceeded in the future.

Acknowledgements

Contributions from Bonnie Hofer (figure preparation) and Jollin Charest (abstract translation) are gratefully acknowledged. Shelley Woods, Alberta Agriculture and Rural Development, estimated the weighted-average irrigation system application efficiency for the irrigation districts in 2011. Robert Riewe, Alberta Agriculture and Rural Development, determined the range of conveyance losses (100 to 130 mm) from estimated gross irrigation district diversion requirements for irrigation based on water conveyance infrastructure characteristics in 2009, and modeled farm-gate demand derived from historical weather data (1928 to 2009).