1. Introduction

Climate includes patterns of temperature, precipitation (rain or snow), humidity, wind and seasons Exquisite impacts on hydrological systems, ecosystems, agriculture and other related systems have been expected (Lal et al., 1998; McCarthy et al., 2001). Longer growing seasons and warmer temperatures may bring benefit in cold regions may be the positive aspects including some negative impacts of reduction of water availability, greater water demand and more frequent extreme weather. These impacts may put agriculture in serious risks (Eitzinger and Kubu, 2009). Kosa and Pongput (2007) reported that, as much as 10% change in production will associated with a change in growing season precipitation by one standard deviation (e.g. Millet in South Asia). Irrigation based on groundwater is adopted to cultivate high-yielding rice varieties and Bangladesh is the world’s fourth biggest rice producing country (Scott and Sharma, 2009; IRRI, 2010). Climate change in Bangladesh is an extremely crucial issue and according to National Geographic (2002), Bangladesh ranks first as the nation most vulnerable to the impacts of Climate Change in the coming decades. The annual rainfall in Bangladesh ranges from 2300 mm (90.6 in) to 2600 mm (102.44 in) but uneven distribution. Ali et al., (2007) found that, the average annual rainfall in Bangladesh in every year was 2486 mm (97.95 inch) and it was about 66.67% of the total year’s rainfall. About 70% to 80% of the total rainfall occurs from month of June to September. Also increased the occurrence of climate-related serious issues like floods, droughts, heat waves and cyclones in future (FAO, 2006; IPCC, 2007; Yu et al., 2010; Ahsan et al., 2011). These changes make a great concern about the drastic consequences on the agricultural crop growth and food security in most of the parts of the world, especially in developing countries (FAO, 2007; IPCC, 2007; WB, 2010; Rouder et al., 2011). Uncertain rainfall pattern creates extreme events such as flood, cyclone, drought resulting adverse effect on human life, animals, eco-system and crop yields especially on rice production (GOB and UNDP, 2009. By 2050, the global water demand of agriculture is estimated to increase by a further 19% due to irrigational needs. Worldwide, agriculture accounts for 70% of all water consumption. The largest sector of Bangladesh’s economy is ‘Agriculture’, which is about 20.29% of the GDP and 47.5% of the labor sector (BBS, 2012). Amount of Total rice production 52,231.00.tons, yield 4.42 tons/hectare, covers 11,820.00 hectare (BBS, 2015). Modern technologies adapt new varieties and irrigation system to fulfil the food security. Impact of climate change is now the main concern for scientist to overcome production losses, with the increasing population. Evapotranspiration (ET) refers to water used by a crop for its growth, tissue building, cooling purposes as well as soil evaporation (Alkaisi and Broner, 2005). ET rates depend upon the factors such as temperature, humidity, solar radiation, wind speed and vegetation characteristics that is transpiring, with significant variation of vegetation types (Allen et al., 1998). The reference crop evapotranspiration (ET_c) which is defined in FAO-24 as “the rate of ET from an extensive surface of 8 to 15 cm tall, green grass cover of uniform height, actively growing, completely shading the ground and not short of water.” If the ET demand exceeds the available water to the plant, then transpiration may stop resulting in crop loss. Therefore, the accurate estimation of ET along with proper knowledge of precipitation and other factors, soil moisture storage capacity can provide measurement of water need for a crop. According to Bangladesh Rice Research Institute (BRRI, 1991), Aman is almost a completely rain-fed rice that grows in the months of monsoon, although it requires supplemental irrigation during planting and sometimes, in the flowering stage, depending on the availability of rainfall. For calculating crop-water demand, irrigation scheduling, preparing input data to any hydrological water-balance models, and assessing and planning regional water resources and evaluating data, among various climatic factors, Reference Crop Evapotranspiration, ET_o is an important key. A 1-mm loss of water through ET across 1 ha crop field is equivalent to 10 m³ (268,000 gallons) of water (Allen et al., 1991). This study is done for identifying the dependable rainfall and water requirement at every stage of aman rice (one varieties) for providing accurate amount of water at every stage in future. When rainfall fails to provide sufficient moisture for normal plant growth, in order to improve and increase yields, supplemental irrigation with little amount of water is applied to plants. This study was carried out to estimate ET_c for two Aman varieties, probability analysis for dependable rainfall and determination of stage wise irrigation for these two Aman rice varieties.

2. Methodology

2.1. Study Area and Crop Data

The study was conducted for Dinajpur district, which was situated in north-west hydrological region. Total area of Dinajpur is 3439.98 sq. km, located between 25010´ and 26004´ north latitudes and in between 88023´ and 89018´ east longitudes. Dinajpur is located in tropical monsoon climate. Rainfall is generally heavy during July and August (BBS, 2009). According to the crop sowing time, 1^st and 2^nd transplanting time was classified. 1^st transplanting time indicates the approximately early sowing time for these two varieties and 2^nd transplanting time indicates the possible late sowing time.

Figure 1. Map of Dinajpur District

Table 1. Name of rice varieties, releasing year, time of transplanting and growing period

Table 2. Length of different growth stages of BR11 and BR22 rice variety

2.2. Data and Methods

Some of the climatic data were found partially missing for the selected study area which collected from Bangladesh Meteorological Department, Dhaka, Bangladesh. Missing data was estimated by following equation (Hydrology, H.M. Raghunath):

(1)

Where, P_x = data of missing station, P₁, P₂….. P_n = data of index stations and n = numbers of index stations.

There are various types of methods used for calculation of reference crop evapotranspiration (ET_o). The Penman-Monteith method is recommended as the sole standard method of determination of ET_o by FAO (Allen et al., 1998). The classic Penman-Monteith method combines both energy and mass balances to model reference evapotranspiration (ET_o) are given below (FAO, 2006):

(2)

Where, Rn=net radiation at the crop surface, MJ/m²/day; G= soil heat flux density, MJ/m²/day;

T= air temperature at 2 m height, °C;

u2= wind speed at 2 m height, m/sec;

es = saturation vapor pressure, kPa;

ea= actual vapor pressure; kPa;

(es - ea)= saturation vapor pressure deficit, kPa;

∆ = slope of vapor pressure curve, kPa/°C and

γ = psychometric constant, kPa/°C.

Crop coefficient (K_c) is a ratio of actual crop evapotranspiration (ET_c) to the reference crop evapotranspiration (ET_o). The growing period of rice has four distinct growth stages: initial, crop development, mid-season and late season. The length of the initial period is highly dependent on the crop, the crop variety, the planting date and the climate.

Allen et al. (1998), K_c for initial stage was considered as 1.05 for both rice varieties.

K_c of the mid-season stage (1.20) was adjusted by considering minimum relative humidity, wind speed and crop height. The equation for adjustment is (FAO, 2006)

(3)

Where, K_cmid =crop coefficient of mid-season stage; U₂ = mean daily wind speed at 2 m height over grass during the late-season growth stage (m/s); RH_min = mean daily minimum relative humidity during the mid-season growth stage (%); and h = the mean plant height during the mid-season stage (m).

The crop coefficient of the development stage of rice was derived by empirical equation considering constant crop coefficient values during the initial and mid-season stages. This empirical equation is given (FAO paper no.56):

(4)

Where, K_cdev = crop coefficient of rice at crop development stage; i is the day number within the growing season (l= length of the growing season); L_stage =length of the stage under consideration (day); ∑L_prev =sum of lengths of all previous stages (day); K_cnext = crop coefficient at the beginning of next stage and K_cprev = crop coefficient at the end of previous stage;

Irrigation was ceased fifteen days before harvest and during the late season stage (Doorenbos et al., 1979). The equation for adjustment is: (FAO, 2006):

(5)

Where, K_cend = crop coefficient of late-season stage; RH_min = mean daily minimum relative humidity during the mid-season growth stage (%); u₂ = mean daily wind speed at 2 m height over grass during the late-season growth stage (m/s) and h= the mean plant height during the mid-season stage (m).

In the crop coefficient approach, actual crop evapotranspiration (ET_c) is calculated by multiplying ET_o by K_c (FAO, 2006):

(6)

Where, ET_crop = actual evapotranspiration (mm/day); K_c = crop coefficient at different growing stages and ET_o = reference crop evapotranspiration at different growing stages of rice (mm/day).

Weibull's ranking method was used for probability analysis of 20 years data (1991-2010). According to this method, yearly rainfall and ET_c data were arranged in descending order. Probability was calculated as follows: (hydrology by H.M. Raghunath)

(7)

Where, N= total number of data and m = order or rank of the observation from the highest value.

The probabilities thus found were plotted on log-normal probability paper. Rainfall and ET_c were plotted against the corresponding probability on the log normal probability paper and the probability curves were drawn. The expected rainfall and ET_c at probability level of 75% was estimated for different growth stages of BR11 and BR22 1991 to 2010 (20 years).

Supplemental irrigation (SI) at different growth stages of rice in aman seasons under Dinajpur districts was calculated by:

(8)

3. Results and Discussions

In table 3, Actual evapotranspiration of BR11 at the 1st transplant time varied from 45 to 83 mm in initial stage, 134 to 180 mm in crop development stage, 120 to 154 mm in mid stage and 36 to 58 mm in late stage. In case of 2^nd transplant time it varied from 43 to 73 mm in initial stage, 133 to 178 mm in crop development stage, 116 to 145 mm in mid stage and 33 to 72 mm in late stage.

Table 3. Actual crop evapotranspiration ETc for BR11

In table 4, Actual crop evapotranspiration of BR22 at the 1st transplant time varied from 40 to 77mm in initial stage, 127 to 179 mm in crop development stage, 116 to 140 mm in mid stage and 33 to 47 mm in late stage. In case of 2^nd transplant time it varied from 41 to 77 mm in initial stage, 126 to 179 mm in crop development stage, 112 to 146 mm in mid stage and 28 to 38 mm in late stage.

Table 4. Actual crop evapotranspiration ETc for BR22

For BR11 (mentioned in table 5), supplemental irrigation was needed during the development stage, mid stage and late stage for 1^st transplanting time and for 2^nd transplanting time, supplemental irrigation needed in mid stage and late stage respectively. Similarly, for BR22 (mentioned in Table 6) supplemental irrigation need was detected as mid stage and late stage for 1^st transplanting time) and for 2^nd transplanting time. In Figure 2(A) and 2(B), data shown was the dependable rainfall at 75% probability level as compared to ET_c for BR11 and BR22 rice variety.

Table 5. Supplemental irrigation for BR11 in Dinajpur District (Dependable Rainfall and ETc at 75% probability level

Table 6. Supplemental irrigation for BR22 in Dinajpur District (Rainfall and ETc at 75% probability level)

Figure 2(A). Dependable rainfall at 75% probability as compared to ETc at initial, development, mid and late stages of BR11 for 1st transplanting and 2nd transplanting in Dinajpur

Figure 2(B). Dependable rainfall at 75% probability as compared to ETc at initial, development, mid and late stages of BR22 for 1st transplanting and 2nd transplanting time in Dinajpur district

4. Conclusions

Major conclusions were drawn on the basis of findings of this study:

i. Adjusted crop coefficients of BR11 and BR22 at different growth stages in different years and sowing times varied due to the changes of relative humidity and wind speed.

ii. Actual crop evapotranspiration of BR11 for 1^st transplant time varied from 359 to 441 mm and 363 to 503 mm in Dinajpur districts. For BR22, actual crop evapotranspiration for 1^st transplanting varied from 361 to 463 mm and 333 to 393 mm for 2^nd transplanting time. Actual crop evapotranspiration varied due to variation of crop coefficient and reference crop evapotranspiration.

iii. For BR11, supplemental irrigation was needed in development, mid and late stage in 1^st transplanting, but for 2^nd transplanting time, supplemental irrigation was needed in mid and late stages. For BR22, supplemental irrigation was needed mainly in mid and late stages for 1^st and 2^nd transplanting.