Printer Friendly

Estimating global climate change impacts on hydropower projects: applications in India, Sri Lanka and Vietnam.

I. Introduction

Due to increasing carbon dioxide carbon dioxide, chemical compound, CO2, a colorless, odorless, tasteless gas that is about one and one-half times as dense as air under ordinary conditions of temperature and pressure.  concentrations it is predicted that global warming global warming, the gradual increase of the temperature of the earth's lower atmosphere as a result of the increase in greenhouse gases since the Industrial Revolution.  would have a variety of environmental and socio-economic impacts over the long term. In recent years particular attention has been paid to hydropower hy·dro·pow·er  
Hydroelectric power.
 generation, because it can offer a supply of renewable energy Renewable energy utilizes natural resources such as sunlight, wind, tides and geothermal heat, which are naturally replenished. Renewable energy technologies range from solar power, wind power, and hydroelectricity to biomass and biofuels for transportation. . At the same time, however, hydropower is clearly among the most vulnerable areas to global warming because water resources are closely linked to climate changes.

To analyze the potential impacts of climate changes on the hydropower industry, the current paper aims to develop a hydrological hy·drol·o·gy  
The scientific study of the properties, distribution, and effects of water on the earth's surface, in the soil and underlying rocks, and in the atmosphere.
 model using a simple multivariate time series technique. The model is applied to three hydropower projects in India, Sri Lanka Sri Lanka (srē läng`kə) [Sinhalese,=resplendent land], formerly Ceylon, ancient Taprobane, officially Democratic Socialist Republic of Sri Lanka, island republic (2005 est. pop.  and Vietnam to check the methodological applicability. The results may be tentative in terms of both methodology and implications, but comparing those three cases, the paper will quantify Quantify - A performance analysis tool from Pure Software.  the major impacts of climate changes on hydrology hydrology, study of water and its properties, including its distribution and movement in and through the land areas of the earth. The hydrologic cycle consists of the passage of water from the oceans into the atmosphere by evaporation and transpiration (or , whence whence  
1. From where; from what place: Whence came this traveler?

2. From what origin or source: Whence comes this splendid feast?

 hydropower projects as a whole. The paper finally attempts to draw some tentative policy implications for hydropower planning and operation.

The world economy is now faced with considerable risk and uncertainty caused by climate changes. Global average temperature would increase by 1.4 to 5.8[degrees]C during the period: 1990-2100. Annual precipitation may also increase or decrease--depending on regions--by 5 to 20 percent, compared with the past 30-year mean (IPCC See IMS Forum. , 2001). Although these figures may have to be considered a preliminary estimate and the debate in fact remains far from conclusive Determinative; beyond dispute or question. That which is conclusive is manifest, clear, or obvious. It is a legal inference made so peremptorily that it cannot be overthrown or contradicted. , it is worth considering potential impacts of global warming on any aspect of the economy and possible mitigation MITIGATION. To make less rigorous or penal.
     2. Crimes are frequently committed under circumstances which are not justifiable nor excusable, yet they show that the offender has been greatly tempted; as, for example, when a starving man steals bread to satisfy

In particular, anticipated climate changes may bring about a dramatically different situation of energy. For instance, if the warming of Earth is accelerating and the decarbonizing energy strategy remains costly, the relative (shadow) price of renewable energy to fossil fuel fossil fuel: see energy, sources of; fuel.
fossil fuel

Any of a class of materials of biologic origin occurring within the Earth's crust that can be used as a source of energy. Fossil fuels include coal, petroleum, and natural gas.
 could be soaring. Some countries with large water resources may become "water-rich" economies, in place of oil-rich countries. But the large upfront investment required for hydropower development may continue posing a challenging question on the fiscal side. The demand for energy may also change given changing climate and mitigation measures. These dynamic manifold manifold

In mathematics, a topological space (see topology) with a family of local coordinate systems related to each other by certain classes of coordinate transformations. Manifolds occur in algebraic geometry, differential equations, and classical dynamics.
 issues have just started attracting keen interest. Shalizi (2007), using a multi-regional global model, simulates energy supply and demand, price trajectory Trajectory

The curve described by a body moving through space, as of a meteor through the atmosphere, a planet around the Sun, a projectile fired from a gun, or a rocket in flight.
 and growth. It is shown that higher energy prices generated by rapid growth in China and India may constrain con·strain  
tr.v. con·strained, con·strain·ing, con·strains
1. To compel by physical, moral, or circumstantial force; oblige: felt constrained to object. See Synonyms at force.

 other countries' growth. The model accounts for investment allocation and future technology but still does not endogenize potential climate changes.

Hydropower projects are one of the areas particularly affected by changes in global and regional climate. This is because hydropower plants have a very long life of more than 50 years. Therefore, the impacts are inevitable even if a significant part of the changes takes place in the distant future. One might think that from the conventional economic point of view, the impact would be presumed small due to discount factors. (1) Nonetheless, this does not mean that the impacts can be ignored, because of the irreversibility Irreversibility
crossing the Rubicon

Caesar passes point of no return into Italy. [Rom. Hist.: Brewer Dictionary, 941]

Humpty Dumpty

all the King’s men failed to reassemble him. [Nuns. Rhyme: Mother Goose, 40]
 of the process as well as social criticality of extreme consequences.

There are three main impacts of climate changes on hydropower projects. First, the available discharge of a river may change, since hydrology is usually related to local weather conditions, such as temperature and precipitation in the catchment area catchment area or drainage basin, area drained by a stream or other body of water. The limits of a given catchment area are the heights of land—often called drainage divides, or watersheds—separating it from neighboring drainage . This will have a direct influence on economic and financial viability of a hydropower project. Moreover, hydropower operations may have to be reconsidered to the extent that hydrological periodicities or seasonality change. The reason is that, if the flow of water changes, different power generating operations, e.g., peak versus base load, would be possible using other designs for water use, such as reservoirs.

Second, an expected increase in climate variability may trigger extreme climate events, i.e., floods and droughts. For instance, a hydrological model indicates a great risk of Bangladeshi suffering from extreme floods, which are led by substantial increases in (mean) peak discharges in the regional three major rivers, Ganges, Brahmaputra and Meghna (Mirza, 2002). One of his scenarios predicts that the volume of water in the Ganges would increase by 5 to 15 percent, depending on changes in temperature.

Finally, closely related to the above, changing hydrology and possible extreme events must of necessity impact sediment sediment, mineral or organic particles that are deposited by the action of wind, water, or glacial ice. These sediments can eventually form sedimentary rocks (see rock).  risks and measures. More sediment, along with other factors such as changed composition of water, could raise the probability that a hydropower project suffers greater exposure to turbine turbine, rotary engine that uses a continuous stream of fluid (gas or liquid) to turn a shaft that can drive machinery.

A water, or hydraulic, turbine is used to drive electric generators in hydroelectric power stations.
 erosion. When a major destruction actually occurs, the cost of recovery would be enormous. An unexpected amount of sediment will also lower turbine and generator efficiency, resulting in a decline in energy generated.

This paper mainly addresses the first issue by analyzing hydrological and weather time series under the assumption of long-run statistical stability. For the purpose of discussing extreme rainfall events and the resulting floods, some other approaches are needed, such as frequency analysis (e.g., Stedinger et al., 1993; Khaliq et al., 2006). A number of probability distribution Probability distribution

A function that describes all the values a random variable can take and the probability associated with each. Also called a probability function.

probability distribution 
 assumptions are tested in this regard (e.g., Malamud and Turcotte, 2006; Bhunya et al., 2007). Even a nonparametric Bayesian method is applicable (O'Connell, 2005). With a regional climate model, Kay et al. (2006) demonstrate the detailed estimates of flood frequency for 15 catchment catch·ment  
1. A catching or collecting of water, especially rainwater.

a. A structure, such as a basin or reservoir, used for collecting or draining water.

 areas in the United Kingdom. In the multivariate context, Samaniego and Bardossy (2007) investigate extreme events in connection with various physiographical and geographic factors. Notably, however, all these models necessitate ne·ces·si·tate  
tr.v. ne·ces·si·tat·ed, ne·ces·si·tat·ing, ne·ces·si·tates
1. To make necessary or unavoidable.

2. To require or compel.
 long hydrological time series, preferably on a daily or hourly basis. To account for the sediment handling issues, even more detailed data elements may be required.

Methodologically, the current paper adopts a simple multivariate stochastic By guesswork; by chance; using or containing random values.

stochastic - probabilistic
 approach, the vector autoregression Vector autoregression (VAR) is an econometric model used to capture the evolution and the interdependencies between multiple time series, generalizing the univariate AR models.  (VAR) model. This seems a natural extension from the conventional univariate time series analysis (e.g., Lettenmaier and Wood, 1993; Salas, 1993; Mohammadi et al., 2006; Wong et al., 2007). Mohammadi et al. (2006) apply an autoregressive moving average (ARMA) model for forecasting a river flow in Iran. Using data on monthly river flows for the past 70 years, it has been found that the ARMA parameters--estimated with "goal programming"--are sufficiently accurate for projection purposes. Wong et al. (2007), in the context of the Saugeen River The Saugeen River is located in southern Ontario, Canada, flowing generally north-west about 160km before exiting into Lake Huron. The river is navigable for some distance, and was once an important barge route. Today the river is best known for its fishing and as a canoe route.  in Canada, propose a more flexible technique, the semi-parametric regression model, which imposes few restrictions on the disturbance DISTURBANCE, torts. A wrong done to an incorporeal hereditament, by hindering or disquieting the owner in the enjoyment of it. Finch. L. 187; 3 Bl. Com. 235; 1 Swift's Dig. 522; Com. Dig. Action upon the case for a disturbance, Pleader, 3 I 6; 1 Serg. & Rawle, 298.  processes.

Our VAR model includes two particularly important variables for hydrological forecasting: temperature and precipitation. There are many other input data that are potentially relevant to hydrology in a river, such as evaporation evaporation, change of a liquid into vapor at any temperature below its boiling point. For example, water, when placed in a shallow open container exposed to air, gradually disappears, evaporating at a rate that depends on the amount of surface exposed, the humidity , soil moisture, catchment characteristics, land use, atmospheric circulation Atmospheric circulation is the large-scale movement of air, and the means (together with the smaller ocean circulation) by which heat is distributed on the surface of the Earth. , polar ice, glaciers This is a list of glaciers.

Due to somewhat sparse information, some glaciers, especially those in the tropics, may no longer exist as listed. This is especially true for glaciers in Africa and New Guinea.
, and even human facilities and activities. Undoubtedly hydrological flows are complex phenomena. In order to keep our model simple and tractable tractable

easy to manage; tolerable.
, however, only temperature and precipitation are selected largely due to data availability Refers to the degree to which data can be instantly accessed. The term is mostly associated with service levels that are set up either by the internal IT organization or that may be guaranteed by a third party datacenter or storage provider. , though there may be a sense that a bias exists in projected temperature and precipitation in many climate models (Bergstrom, 2001; World Bank, 2007).

The VAR model seems appropriate to examine hydrological data for three reasons. First, it is expected to uncover dynamic interactions among the variables of interest. The flow of a river is affected by regional climate, and vice versa VICE VERSA. On the contrary; on opposite sides. . Both belong to the world water cycle. Second, hydrologic time series tends to exhibit a set of statistical properties necessary for VAR, such as stationarity (Salas, 1993; Koutsoyiannis, 2006). The fundamental assumption for estimating a VAR model is that the first two moments exist and are covariance Covariance

A measure of the degree to which returns on two risky assets move in tandem. A positive covariance means that asset returns move together. A negative covariance means returns vary inversely.
 stationary. Finally, it is computationally com·pu·ta·tion  
a. The act or process of computing.

b. A method of computing.

2. The result of computing.

3. The act of operating a computer.
 easy to perform.

The paper is organized as follows. Section II provides an overview of the three hydropower projects analyzed an·a·lyze  
tr.v. an·a·lyzed, an·a·lyz·ing, an·a·lyz·es
1. To examine methodically by separating into parts and studying their interrelations.

2. Chemistry To make a chemical analysis of.

 in the following sections. Section III describes the empirical model and data issues. Section IV presents the main estimation estimation

In mathematics, use of a function or formula to derive a solution or make a prediction. Unlike approximation, it has precise connotations. In statistics, for example, it connotes the careful selection and testing of a function called an estimator.
 results. Demonstrating the climate change impacts calibrated cal·i·brate  
tr.v. cal·i·brat·ed, cal·i·brat·ing, cal·i·brates
1. To check, adjust, or determine by comparison with a standard (the graduations of a quantitative measuring instrument):
 from the hydrological models, Section V discusses some policy implications for preparing and operating hydropower projects.

II. Overview of Three Hydropower Projects

Despite a number of global and regional climate models, it is still open to argument what would be the most realistic climate change scenario. Envisaged climate changes vary depending on underlying assumptions, such as natural and eco-systems, economic growth, technology, and population growth. The Intergovernmental Panel on Climate Change “IPCC” redirects here. For other uses, see IPCC (disambiguation).
The Intergovernmental Panel on Climate Change (IPCC) was established in 1988 by two United Nations organizations, the World Meteorological Organization (WMO) and the United Nations Environment
 (IPCC) originally provides seven scenarios, and other organizations are adding more. For instance, a scenario "A1a" of the CSIRO CSIRO Commonwealth Scientific & Industrial Research Organization (Australia)  Atmospheric Research Atmospheric Research (ISSN 0169-8095) is scientific journal dealing with the part of the atmosphere where meteorological events occur; intended for atmospheric scientists (such as meteorologists and climatologists), aerosol scientists, and hydrologists.  model, which the following analysis will partially rely on, assumes rapid economic growth and global integration of economies. The IPCC database generates mean temperature ([degrees]C) and daily precipitation (mm) in the 2020s and 2050s, which represent the periods: 2010-2039 and 2040-2069, respectively (Figure 1). Again, it cannot be overemphasized that this is merely one of the stories; the climate forecasts substantially differ from scenario to scenario.


Three hydropower projects are used for case studies; these are located in India, Sri Lanka and Vietnam. Table 1 summarizes their major characteristics. While the Vishnugad Pipalkoti Hydro Electric Project (VPHEP) in India is still in preparation, the other two, the Upper Kotmale Hydro Power Project (UKHPP) in Sri Lanka and the Thac Mo Hydropower Station Extension Project (TMHSEP) in Vietnam, have already been launched.

The selected projects look very different in various aspects. The Vishnugad Pipalkoti project will be a typical "run-of-river" station though it has a small storage capacity for diurnal diurnal /di·ur·nal/ (di-er´nal) pertaining to or occurring during the daytime, or period of light.

1. Having a 24-hour period or cycle; daily.

 variations. On the other hand, the Upper Kotmale project includes a daily reservoir of 800,000 m3, which allows it to operate for peak load power generation. Correspondingly, its plant utilization rate is assumed 40 percent. The Thac Mo project aims to extend the existing generation capacity of 150 MW and relies for water resources on the existing large Thac Mo Reservoir, which lies about 100 km north of Ho Chi Minh City Ho Chi Minh City, formerly Saigon, city (1997 pop. 5,250,000), on the right bank of the Saigon River, a tributary of the Dong Nai, Vietnam. .

How to operate a hydropower plant is largely dependent on the reservoir capacity. VPHEP is expected to contribute to providing base load energy during the rainy rain·y  
adj. rain·i·er, rain·i·est
Characterized by, full of, or bringing rain.

raini·ness n.

 season, which is consistent with the basic fact that Northern India still lacks electricity in absolute terms (Alg.) such as are known, or which do not contain the unknown quantity.

See also: Absolute
, possibly owing to owing to
Because of; on account of: I couldn't attend, owing to illness.

owing to prepdebido a, por causa de 
 the recent economic buoyancy buoyancy (boi`ənsē, b`yən–), upward force exerted by a fluid on any body immersed in it. Buoyant force can be explained in terms of Archimedes' principle. . In the dry season it may supply peak load energy using its attached hourly storage. The other two projects basically aim to operate on a peak load basis. In Sri Lanka, the Upper Kotmale project is one of the last large-scale hydropower projects. Most water resources have already been utilized in this country. The extended part of the Thac Mo hydropower station will provide energy to meet the residual demand that the existing plant cannot supply.

The three projects also differ in size and cost. While the VPHEP is intended to generate about 1,800 GWh of energy per annum Per annum

, the much smaller amount of energy will be delivered by the UKHPP and TMHSEP. The Upper Kotmale project may cost about 300 million U.S. dollars, but the total cost of the Thac Mo project was estimated at 50 million U.S. dollars because it is a relatively small extension work and does not include any new storage construction.

Climate changes are likely to occur differently among project locations. According to according to
1. As stated or indicated by; on the authority of: according to historians.

2. In keeping with: according to instructions.

 the IPCC forecasts, around the project area of Northern India temperature is expected to increase by about 1-2[degrees]C in the next two decades (Table 2). In the fall and early winter, a warming may be relatively modest on the order of 1/2-1[degrees]C. Toward the 2050s an additional warming of about 2[degrees]C would be expected in this region. On the other hands, monthly precipitation would decrease by about 1/2-1 mm per day in most months but might upsurge in the middle of the rainy season. As a result, precipitation would concentrate even more on the rainy season.

In Sri Lanka temperature is likely to increase evenly all the year around. Precipitation would concentrate on one of the current double-humps; more rainfall is expected from June to September, but the dry season would have less precipitation. Around the Thac Mo project site in Southern Vietnam, temperature would rise by 1/2-1 percent, especially in winter. The area will experience increases in rainfall in most months, but the increments may be relatively small, compared with Northern India and Sri Lanka.

III. Model and Data

Following the existing literature (e.g., Salas, 1993; Mohammadi et al., 2006), a simple multivariate stochastic model, VAR, is considered. Stationarity is a key requirement for performing the VAR model. Hydrologic data defined on an annual time scale are generally characterized char·ac·ter·ize  
tr.v. character·ized, character·iz·ing, character·iz·es
1. To describe the qualities or peculiarities of: characterized the warden as ruthless.

 stationary unless there are large-scale climate variability, natural disruptions and human-induced changes such as reservoir construction (Salas, 1993). A typical hydrologic process is composed of three parts: (i) a deterministic 1. (probability) deterministic - Describes a system whose time evolution can be predicted exactly.

Contrast probabilistic.
2. (algorithm) deterministic - Describes an algorithm in which the correct next step depends only on the current state.
 part resulting from natural physical periodicities, (ii) an aperiodic a·pe·ri·od·ic
Not occurring periodically.

 deterministic part which is often referred to as trends, and (iii) a stationary random component. Once detrending and adjusting seasonarity, the remaining time series tends to be covariance stationary (Koutsoyiannis, 2006).

Our hydrological data seem to have strong seasonal regularities on an annual basis (Figure 2). Outstanding hikes in water flow have been observed twice a year in the Sri Lanka's case and once a year in the rest of the cases. Another intuitive finding from the figure is that while a water flow in the Kotmale River appears to have declined over the past five decades, the Thac Mo discharge may have had an increasing steady tendency in recent years. The feasibility sturdy sturdy

neurological disease in sheep caused by the pressure of a Taenia multiceps metacestode. Called also gid.
 recognizes the fact that the discharge at the Thac Mo Reservoir has increased since the construction of the original dam. The Alaknanda River The Alaknanda River is a tributary of the Ganges River that begins at the confluence of the Satopanth and Bhagirath Kharak glaciers in Uttarakhand. It merges with the Bhagirathi river near Devprayag after flowing for approx. 229 km through the Alaknanda valley.  does not seem to have a significant time trend component.


In addition to hydrology, both precipitation and temperature series are also apt to exhibit considerable seasonality. However, it is common that the former has more irregularities than the latter (Figure 3). (2)


More formally, it has been found that in the Vishnugad Pipalkoti case, hydrological series looks similar to the correlogram of an autoregressive multiplicative seasonal series, (1-[alpha]L)(1-[phi][L.sup.12])[x.sub.1], where [x.sub.1] is a random variable at period t. L denotes the lag operator In time series analysis, the lag operator or backshift operator operates on an element of a time series to produce the previous element. For example, given some time series

. When applying correlogram, which is a plot of the autocorrelation Autocorrelation

The correlation of a variable with itself over successive time intervals. Sometimes called serial correlation.
 coefficient coefficient /co·ef·fi·cient/ (ko?ah-fish´int)
1. an expression of the change or effect produced by variation in certain factors, or of the ratio between two different quantities.

 as a function of the number of lags included in the assumed first-order autoregressive (AR(1)) process, the autocorrelation coefficients first decline and then increase again to a peak in a 12-month cycle. The partial autocorrelations have a positive and negative spike at 1 and 13 lags, respectively (Figure 4). These features are reasonable because hydrology and weather data would likely have a 12-month cycle with some white noise and display some continuity over adjacent months (Johnston and DiNard, 1997). The other two cases, the Kotmale River and the Thac Mo Reservoir, show the same pattern of correlograms.


When removing monthly periodicities in a deterministic manner, hydrologic and weather series indeed come to exhibit clearer stationarity. Figure 5 depicts the seasonally adjusted Seasonally adjusted

Mathematically adjusted by moderating a macroeconomic indicator (e.g., oil prices/imports) so that relative comparisons can be drawn from month to month all year.
 time series of the Vishnugad Pipalkoti case, for example. Some extreme observations--compared with the long-term monthly mean--remain striking; however, the seasonality disappears. Of particular note, in this case the seasonally adjusted precipitation series may have a declining time trend especially in the past ten years, while no clear trend can be detected graphically in hydrology and temperature series. (3)


The augmented Dickey-Fuller test In statistics and econometrics, an augmented Dickey-Fuller test (ADF) is a test for a unit root in a time series sample. It is an augmented version of the Dickey-Fuller test to accommodate some forms of serial correlation.  with 13 lags is performed to test for stationary. The number of lags is selected following the above argument. As shown in Table 3, all variables in the three cases, but one, have been found stationary in both level and first-difference terms. At least the unit root hypothesis is strongly rejected at the conventional significance level. The hydrology data associated with the Thac Mo Reservoir is exceptional. It may potentially exhibit a unit root; but with the seasonally adjusted series the null hypothesis null hypothesis,
n theoretical assumption that a given therapy will have results not statistically different from another treatment.

null hypothesis,
 can be easily rejected (ADF (1) (Application Development Facility) An IBM programmer-oriented mainframe application generator that runs under IMS.

(2) (Automatic Document Feeder) A paper stacker that feeds one sheet of paper at a time into the unit.
 test statistics is estimated at -10.819). Additional attention will be paid to this unit root case in the following analysis.

The above discussion allows us to apply the VAR model with deterministic periodicity periodicity /pe·ri·o·dic·i·ty/ (per?e-ah-dis´i-te) recurrence at regular intervals of time.

 and trend components. Consider the following system of equations with three variable, i.e., ln HYDRO, ln TEMP, and ln PREC :

[z.sub.t] = v+[[DELTA].sub.p][A.sub.p][z.sub.t-p]+[c.sub.month]+[rho]t+[u.sub.t]

where [z.sub.t] is a 3x1 random vector, [A.sub.p] is a 3 x 3 matrix of parameters to be estimated, and [c.sub.month] is a set of monthly dummy Sham; make-believe; pretended; imitation. Person who serves in place of another, or who serves until the proper person is named or available to take his place (e.g., dummy corporate directors; dummy owners of real estate).  variables. In addition, [rho] is a linear detrending parameter (1) Any value passed to a program by the user or by another program in order to customize the program for a particular purpose. A parameter may be anything; for example, a file name, a coordinate, a range of values, a money amount or a code of some kind. .

An alternative way of estimating the hydrological model may be to first remove seasonality and trends and then analyze detrended series (e.g., Salas, 1993). However, the current model with deterministic seasonality included in the system is expected to yield a more efficient estimate.

HYDRO is defined as the monthly discharge (in million cubic meters ([m.sup.3])) of each river at the project site. The data come from each of feasibility studies. Temperature and precipitation data depend on the NOAA NOAA
National Oceanic and Atmospheric Administration

Noun 1. NOAA - an agency in the Department of Commerce that maps the oceans and conserves their living resources; predicts changes to the earth's environment;
 GHCN GHCN Global Historical Climatology Network  Monthly Database Version 2; the data from the observatory closest to the project location are used. (4) In the NOAA database, however, there is no available comprehensive data for Southern Vietnam. Alternatively, the observed regional weather time series provided by the IPCC Data Distribution Centre (DDC See VESA DDC. ) are borrowed. (5) But unlike the NOAA database, these time series are available only up to 1990, and may have poor spatial representation. Thus, when using them, we may risk underestimating the most recent impacts of global warming. Whereas TEMP is defined as the monthly average temperature converted to Fahrenheit in order to avoid taking logarithms of negative numbers, PREC is total monthly rainfall measured in millimeters (mm). (6) When there is no precipitation in a particular month, it is set to a very small positive number, but not zero.

How are these three variables related to each other? Table 4 shows simple correlations. The extent to which hydrology is linked to climate differs among project locations. Precipitation may be most relevant to hydrological series. However, temperature may be positively or negatively associated with river flow. There is no serious multicollineality problem in our data.

An important specification question in using the VAR-type model is how many lags should be included in the system. When removing periodicities on a unilateral unilateral /uni·lat·er·al/ (-lat´er-al) affecting only one side.

On, having, or confined to only one side.
 variable basis, the correlograms of seasonally adjusted hydrologic series may suggest that in the Vishnugad Pipalkoti case, for instance, the first three or four orders might be autocorrelated (Figure 6). Formally, the Akaike's information criterion There are a number of statistics that can act as an information criterion. They include:
  • Akaike's information criterion
  • the Bayesian information criterion, also known as the Schwarz information criterion
  • Hannan-Quinn information criterion
 (AIC) lag-order selection statistics is estimated (Akaike, 1973). As shown in Table 5, the maximum number of lags to be included in the model is 2 and 3 for the Vishnugad Pipalkoti and Upper Kotmale cases, respectively. In the Thac Mo hydro case, only one lag may need including. These results are robust, regardless of whether or not a time trend component is introduced in the model.


IV. Estimation Results

Estimated hydrology equation

Four VAR models are performed for each case. Table 6 shows the results of the Vishnugad Pipalkoti case. Only the hydrology equation in question is reported; the other two equations are omitted, though jointly estimated. It is found that hydrology would be reduced by higher temperature and increased by more precipitation. One might think that it is intuitively reasonable, because under the global warming scenario, some regions would suffer from chronic droughts and heat waves simultaneously. Importantly, however, it is noteworthy that the results presented in this paper cannot be overgeneralized. For instance, it is another likely story that temperature is positively correlated cor·re·late  
v. cor·re·lat·ed, cor·re·lat·ing, cor·re·lates
1. To put or bring into causal, complementary, parallel, or reciprocal relation.

 with hydrology; some tropical areas may come to have frequent downpours under high-temperature circumstances. But this is not the case in the VPHEP area.

It is also found that the rapid flow season--which can statistically be defined as March to August according to the estimated monthly coefficients--has a systematically different hydrological flow from our baseline month, i.e., January. Particularly in May, June and July, the Alaknanda River exhibits strong seasonality. On the other hand, the hydrological series does not seem to be trending, even if a trend component is introduced in the equations. The difference is minimal between the models with and without trends.

When a set of zero restrictions are imposed on the coefficients which are not significant in the unrestricted models, the significance of the parameters in the system generally improves while the key results remain unchanged. The restricted models are more reliable in the sense that they have insignificant autocorrelation in the residuals and lower skewness Skewness

A statistical term used to describe a situation's asymmetry in relation to a normal distribution.

A positive skew describes a distribution favoring the right tail, whereas a negative skew describes a distribution favoring the left tail.
. The hypothesis of no autocorrelation cannot be rejected in the restricted models, and the normality normality, in chemistry: see concentration.  hypothesis will be accepted at the 1 percent significance level, though rejected at the 5 percent level. Because of the maximum modulus See modulo.  of the eigenvalues eigenvalues

statistical term meaning latent root.
 being less than one, all the eigenvalues in the system lie inside the unit circle. This means that the estimated system is fairly stable.

Tables 7 and 8 show the VAR estimates of the hydrological equation for the Upper Kotmale and Thac Mo cases, respectively. Similar to the above, the hydrological discharge in the Kotmale River would increase with rainfall and decrease with temperature. In this case, however, temperature seems to have a much more powerful effect than precipitation. It is also shown that the deterministic rainy season appears much long--from April to December--at least when inferring from hydrology.

By contrast, the monthly discharge at the Thac Mo Reservoir may be less related to temperature and precipitation in a statistical sense; both coefficients are insignificant. Rather, the river flow is determined by only its own lagged values and deterministic seasonality. Interestingly, in addition, the Ba River may have a positive time trend, meaning that the amount of available water would become larger as time rolls on, in spite of seasonality and stochastic changes.

Long-term time trends

From the long-term perspective, it is of particular interest whether the system involves a time trend component. Table 9 summarizes the trend coefficients in each unrestricted model. As touched upon in the above, only the hydrological series at the Thac Mo Reservoir has a positive and significant time trend among our sample projects.

In all cases, however, temperature series are highly likely to include a positive trend component. This means that these three project areas would experience certain global warming anyway. The coefficients are very small in scale but still statistically significant. It can be interpreted to mean that the Vishnugad Pipalkoti area would undergo a 0.01[degrees]C increase in temperature per decade, no matter what one would do. Similarly, the increases in temperature over a decade are estimated at 0.003[degrees]C and 0.009[degrees]C in the Upper Kotmale and Thac Mo project areas. (7) Meanwhile, the negative coefficient in the precipitation equation of the Vishnugad Pipalkoti case consistent with our prior expectation in Figure 5. Rainfall in this area would likely continue declining in the very long run.

Impulse response In simple terms, the impulse response of a system is its output when presented with a very brief signal, an impulse. While an impulse is a difficult concept to imagine, and an impossible thing in reality, it represents the limit case of a pulse made infinitely short in time  function

The impulse response function is computed to see the short-term impacts within the system (Figure 7). It illustrates the effect of a one-standard error shock originating from a variable in the system on other endogenous endogenous /en·dog·e·nous/ (en-doj´e-nus) produced within or caused by factors within the organism.

1. Originating or produced within an organism, tissue, or cell.
 variables through the estimated dynamic structure. Note that we use the restricted models, and thus the cases where an impulse does not have a direct impact on the dependent variable are omitted from the figure. For the Vishnugad Pipalkoti project, an instantaneous in·stan·ta·ne·ous  
1. Occurring or completed without perceptible delay: Relief was instantaneous.

 increase in temperature would likely reduce the following hydrological series. On the other hand, increasing rainfall leads to a higher level of discharge.

The system seems to have a relatively long adjustment process; a given shock will disappear after more than one year.

The Kotmale River hydrology follows the same story. If it rains more, the water flow would increase. If it is hotter than usual, then the river tends to be lean. In the case of the Thac Mo hydropower project, the restricted models have no direct impact of climate changes on hydrology. Accordingly, these impulse response functions are not presented.


Causality causality, in philosophy, the relationship between cause and effect. A distinction is often made between a cause that produces something new (e.g., a moth from a caterpillar) and one that produces a change in an existing substance (e.g.  

Causality tests are one of the advantages of estimating a multivariate stochastic system; this feature cannot be used when analyzing only a univariate hydrological time series or imposing the meteorological me·te·or·ol·o·gy  
The science that deals with the phenomena of the atmosphere, especially weather and weather conditions.

[French météorologie, from Greek
 or physical hydrologic relationship in advance. The conventional Granger causality Granger causality is a technique for determining whether one time series is useful in forecasting another. Ordinarily, regressions reflect "mere" correlations, but Clive Granger, who won a Nobel Prize in Economics, argued that there is an interpretation of a set of tests as  test can reveal what causes what. In the Vishnugad Pipalkoti case, temperature Granger-causes water flow of the Alaknanda River (Table 10). The null hypothesis that temperature is irrelevant in the hydro equation can be rejected at the 10 percent significance level, and the hypothesis of hydrology being unimportant un·im·por·tant  
Not important; petty.

unim·portance n.
 to determine temperature cannot be rejected. In the same manner, rainfall also Granger-causes the river flow.

By contrast, at the Kotmale River, temperature seems to cause the water level, and vice versa. The null hypothesis that temperature influences hydrology cannot be rejected, but at the same time, the hypothesis of hydrological series being critical in the temperature equation can also be accepted. Statistically, rainfall does not cause hydrology.

Finally, in the Thac Mo case there is no conclusive causal relationship between climate and hydrology, though temperature Granger-causes precipitation in the region. In sum, hydrological series are likely to be impacted on by climate changes, particularly temperature. However, it may vary on a case-by-case basis.

V. Discussion

Hydrological forecasts

What does the above mean from a hydropower project perspective? First, it means that future hydrological series may be different from what one envisages at the project preparation stage. In ex ante assessing a hydropower project, it is broadly common that the 90 percent dependable hydrological level--which is referred to as a baseline hereinafter--is used as a forecast of water flow available in the future. It is a very conservative approach, which is high-principled for cautious project preparation purposes. However, it is worth recalling that this is just one of the univariate nonparametric point estimates, leaving most hydrological information unused. (8)

As shown in Figure 8, the hydrological forecasts in 2025 based on our empirical results in fact look very different from data in the conventional 90 percent dependable year. The figure includes two types of forecasts: One is the dynamic forecasts, which are calibrated from the last observation in the sample, following the estimated system of equations. The other is one-step-ahead projections, which are calculated by fitting a set of values for the estimated equations to obtain the predicted values in the next period. In this regard the IPCC forecasts presented in Table 2 are adopted as a reference point. (9)

In both Vishnugad Pipalkoti and Thac Mo cases, the rainy season would have higher levels of water than the baselines. (10) However, in the lean season water resources may become even more limited. For the Upper Kotmale project, the river would have a greater flow of water almost all the year around. Our dynamic forecasts look more prone to be greater than the one-step-ahead projections. The reason is that the long-run calibration calibration /cal·i·bra·tion/ (kal?i-bra´shun) determination of the accuracy of an instrument, usually by measurement of its variation from a standard, to ascertain necessary correction factors.  is very sensitive to changes in parameter estimates; a small change in the coefficients could yield a very different picture of the future. Notably, however, the difference between the dynamic forecasts and one-step-ahead projections is relatively tolerable tol·er·a·ble  
1. Capable of being tolerated; endurable.

2. Fairly good; passable. See Synonyms at average.

 in any time series.


Impacts on power generation

To focus on the effect of climate changes while controlling for the existing difference in project design, objective and operation, suppose that all plants provide base load energy, meaning that they always operate as long as water resources are available. The individual physical characteristics of plants hold constant, such as installed capacity and net head. The baseline scenario also assumes that there is no major storage capacity. (11)

Without large storage capacity, just like the Vishnugad Pipalkoti project, the implication of changing hydrological series is direct. If the level of river flow is above the maximum design water discharge of the power station, there is no impact at least in terms of the amount of energy generated. (12) For instance, the design discharge of VPHEP is 224 [m.sup.3] per second or about 580 million [m.sup.3] per month. Thus, the possible large increase in hydrology from June to August in 2025 could not be exploited for power generation purposes. Rather, in the lean season the power station would likely be faced with a severer water constraint Constraint

A restriction on the natural degrees of freedom of a system. If n and m are the numbers of the natural and actual degrees of freedom, the difference n - m is the number of constraints.

Table 11 shows the predicted impacts of climate changes on electricity generation. The amount of energy is calculated on a monthly basis in each scenario, and annual energy is the summation summation n. the final argument of an attorney at the close of a trial in which he/she attempts to convince the judge and/or jury of the virtues of the client's case. (See: closing argument)  of monthly values. Due to increased river discharges during the early and late rainy season, the Vishnugad Pipalkoti power station would be able to generate more energy in 2025. However, this increment To add a number to another number. Incrementing a counter means adding 1 to its current value.  may be relatively modest at about 7.5 percent. The impact of decreased water in the lean flow season would be marginal, because the baseline scenario has already taken into account the fact that the Alaknanda River even now has the very low level of water flow during the lean season. Because of lack of sufficient storage capacity, the Vishnugad Pipalkoti power station will not fully take advantage of water resources available in the high-water season.

In the Upper Kotmale case, a projected increase in water flow may allow to generate about 45 percent more energy than the baseline level. Significantly, this is because the installed capacity is large enough to absorb increasing water flow. Still, the load factor will be estimated at about 40 percent. Although the large installed capacity of the Upper Kotmale hydropower station is intended to supply peak energy given storage for a few days, it might have the additional advantage of exploiting increased hydrological resources for power generation. (13) It will also exhibit certain resistance to increased variability in water flow and extreme events.

Provided that the large-scale reservoir is not used, annual energy generated by the Thac Mo power station might decline given the 2025 hydrological dynamic forecasts. The expected water flow has a large volatility and cannot be absorbed by its relatively small installed capacity of 75 MW. The negative impact of lower water levels in the dry season would be dominant in this case.

In reality, however, these hydro stations have storage capacities to a greater or lesser extent. Only the Thac Mo project has a large-scale reservoir of 1,250 million [m.sup.3] for about six months. Recall that the station is intended to supply peak load energy. The Upper Kotmale, which also aims at providing three-hour peak energy, has only a daily storage. The Vishnugad Pipalkoti project has an hourly storage.

The benefit from a large-scale reservoir is apparent in our monthly-based analytical framework. Under the assumption of the maximum use of the existing storage capacity, the Thac Mo hydropower station would be able to increase to 524 GWh from 383 GWh of the baseline case. (14) This implies that having a storage capacity is useful for accommodating increased seasonality in hydrological series. The Vishnugad Pipalkoti project does not benefit from such seasonal variation adjustment, because its storage capacity is small. The Upper Kotmale hydropower plant does not benefit either; but this is because its installed capacity is large enough to use all the flow of the Kotmale River, even if it increases.

Impacts on project viability assessment

As annual energy changes, the economic and financial project viability might also change. Table 12 presents the internal rate of return (IRR IRR

In currencies, this is the abbreviation for the Iranian Rial.

The currency market, also known as the Foreign Exchange market, is the largest financial market in the world, with a daily average volume of over US $1 trillion.
) corresponding to each scenario. For simplicity, it is assumed that the project cost is distributed evenly for the first five years before the following 30-year operation. Annual operation and maintenance costs are set at 1.5 percent of total project costs. The price (or benefit) of energy generated is assumed 7 U.S. cents per kWh in all cases, despite the fact that it varies across countries and across types of customers. This is just for comparison purposes. It is worth noting that peak load energy should be estimated to be economically more valuable in reality. No other economic benefits and costs are accounted for. Finally, the climate change scenario assumes that the baseline hydro energy is used for the first 10 years of operation, and the estimated dynamic forecasts are applied afterwards af·ter·ward   also af·ter·wards
At a later time; subsequently.

afterwards or afterward

later [Old English æfterweard]

Adv. 1.

The economic effect of changes in energy generated has been found relatively small contrary to prior expectations. This is mainly attributable to our assumption that climate changes would realize 10 years after the power station commissioning. For the Vishnugad Pipalkoti project, climate changes would increase the IRR by only 0.3 percentage points. In the Thac Mo case, a changing climate might lower the IRR, but with its attached reservoir the rate of return would increase from 28.8 percent to 29.6 percent. However, provided that climate changes affect hydrology from the beginning of the plant operation, the rate of return to the Thac Mo project would rise to about 35 percent. This may indicate a pitfall pit·fall  
1. An unapparent source of trouble or danger; a hidden hazard: "potential pitfalls stemming from their optimistic inflation assumptions" New York Times.
 in assessing a hydropower project in economic terms; the future climate change impacts are generally underestimated in the IRR calculation, even though they are environmentally and socially significant.

In the Upper Kotmale case, a substantial increase in electricity production could be expected because of its margin of installed capacity, resulting in a higher IRR of 6.4 percent. The existing daily storage does not directly affect this result. These pieces of evidence suggest that having larger installed capacity and some storage capacity might be well worth consideration.

Importantly, the above discussion does ignore the likely implication on the cost side. Larger installed capacity must of necessity bring about higher construction costs. If a large-scale storage capacity is planned, the additional costs would be enormous not only financially but also socially. For example, consider a 50 percent increase in total project costs for constructing a six month storage capacity. The optimal size of reservoir ranges from several hours to over a year, depending on geological and environmental conditions and operational objectives. The storage capacity selected here is roughly equivalent to the gross storage at the full reservoir level of the Thac Mo Reservoir. Note that in the Thac Mo case the project cost does not include any fraction of past investment in the original Thac Mo Reservoir; thus, this additional cost scenario may still be meaningful even in the Thac Mo case.

As shown in the last two columns of Table 12, the project viability is much more sensitive to the presumed cost increase rather than changing hydrological flows. An additional investment cost would dramatically lower the IRRs. For instance, the rate of return for the Vishnugad Pipalkoti project drops by 5 percentage points under the baseline assumption.

Given our estimated hydrological forecasts, the project viability could improve to a certain extent, thanks to the leveled hydrological series by the additional storage capacity. However, such benefits may not be fully justifiable jus·ti·fi·a·ble  
Having sufficient grounds for justification; possible to justify: justifiable resentment.

 from the IRR perspective. It depends on the cost. Notably, in fact, such a large-scale reservoir has been found overinvestment in the Vishnugad Pipalkoti case; probably a 2-3 month reservoir might be sufficient to follow our hypothetical Hypothetical is an adjective, meaning of or pertaining to a hypothesis. See:
  • Hypothesis
  • Hypothetical
  • Hypothetical (album)
 operating rule of the storage. Thus, larger generation and storage capacities may be a measure against uncertain climate changes; but these options may be expensive, and the potential environmental and social costs could also be considerable. (15) A broad and consistent evaluation will be needed for further assessment.

Limitation of the model

The above discussion has several limitations. First of all, the hydrological projections might be underestimated, because they are essentially estimated based on the past climate and hydrological time series. Including more time series observation contributes to improving statistical reliability but risks underestimating the recent trend in river runoff Runoff

The procedure of printing the end-of-day prices for every stock on an exchange onto ticker tape.

If the "tape is late" then it can take a long time to print off all the closing prices.
 and climate variables.

Second, despite their potential significance, the impact of extreme events is not captured in the above model because of both data and methodological limitations. As shown in Figure 5, for example, the Alaknanda hydrological series appears to have become more volatile in recent years. However, the analysis based on monthly data cannot explain extreme events, such as flash floods and rain floods caused by extremely heavy precipitation in a few days. (16) Any econometric e·con·o·met·rics  
n. (used with a sing. verb)
Application of mathematical and statistical techniques to economics in the study of problems, the analysis of data, and the development and testing of theories and models.
 technique is more or less designed to measure an average effect, ignoring outliers like floods and droughts.

Notably, though, the above projections are still considered suggestive sug·ges·tive  
a. Tending to suggest; evocative: artifacts suggestive of an ancient society.

. For example, it can be shown that in the Vishnugad Pipalkoti case the skewness of a monthly water flow distribution is likely to increase from 0.74 to 1.02 by the 2020s. Generally, Figure 8 also indicates a considerable increase in water flows particularly during the rainy season in all cases. Especially for the Vishnugad Pipalkoti hydropower project, in comparison with the designed capacity the projected surge in river discharge may not be ignorable from the point of view of flood and sediment risk management.

Third, one might question whether the stochastic model is generally suitable for this type of study. It is open to discussion. The above analysis finds the hydrological series at the Thac Mo Reservoir may exhibit a unit root and thus not be applicable to the VAR technique. (17) When detrended data are used, the result has been found quite similar to the result presented above. Nonetheless, even if the model is appropriate, there is another level of problem. For example, Wong et al. (2007) claims that the assumption of a river flow linearly depending on its lagged values is questionable. Our VAR model does not rely on the linearity restriction, but it is simply a log-linear model, which still imposes certain restrictions on the system, e.g., constant elasticity.

In addition, some variables that are critically related to hydrological flows may be omitted from our model. Bergstrom et al. (2001) point out that a key to successful hydrological modeling is to properly account for soil moisture, which is not included in the above model. There are possibly other omitted variables. In the context of Northern India, for instance, the retreating Himalayan glaciers may have to be taken into consideration. (18) However, there may be a tradeoff; more variables will involve more uncertainty. Particularly, evapotranspiration evapotranspiration

Loss of water from the soil both by evaporation from the soil surface and by transpiration from the leaves of the plants growing on it. Factors that affect the rate of evapotranspiration include the amount of solar radiation, atmospheric vapor pressure,
 in a future clime may involve serious uncertainty to be modeled (Bergstrom et al, 2001). Mountain areas where hydropower projects are often located may be more complicated because mountains are among the most fragile environments; the rate of warming in mountain systems is expected to be two to three times higher than that recorded during the 20th century (Nogues-Bravo et al., forthcoming).

There is a piece of evidence to support the validity of the used stochastic model. When comparing our dynamic forecasts of temperature and precipitation with the IPCC projections, temperature forecasts seem well comparable (Figure 9). On the other hand, precipitation forecasts are broadly consistent but may be underestimated in some cases, such as the Vishnugad Pipalkoti project. The difference may be attributed to the factors that the IPCC model accounts for and our VAR model does not. Obviously, again, the IPCC projections, as such, may have to be interpreted with some caution.


Finally, the climate forecasts may need refining refining, any of various processes for separating impurities from crude or semifinished materials. It includes the finer processes of metallurgy, the fractional distillation of petroleum into its commercial products, and the purifying of cane, beet, and maple sugar  furthermore in terms of spatial representation. The above discussion partly depends on the global climate change model for one-step-ahead climate projections. Ideally, however, the basin-level climate forecasts would be more appropriate input (e.g., Kothyari and Singh, 1996; Pant and Kumar, 1997; Lal and Aggarwal, 2001). The above analytical framework implicitly assumes that temperature and precipitation data at or close to the project site could represent all climate conditions over the upper basin--e.g., surface water, groundwater and soil moisture--that would result in flow at the dam location. This may not always hold. At the same time, however, if all the relationship were modeled in a physical hydrologic manner, there would be no room where the statistical hydrology approach could be performed, because there would be no degree of freedom. In other words Adv. 1. in other words - otherwise stated; "in other words, we are broke"
put differently
, a missing gap that should be bridged is, to my best knowledge, the comparison between those two approaches. A further research will be needed to answer whether or not they are compatible and which is better if not.

VI. Conclusion

The world economy is now faced with considerable risk and uncertainty caused by climate changes. Increasing attention has been paid to hydropower generation in recent years, because it is renewable energy. However, hydropower is one of the industries that would be most likely to be affected by changes in global and regional climate. The paper applies a hydrological model using a VAR technique to three hydropower projects in India, Sri Lanka and Vietnam.

The possible climate change impacts have rarely been evaluated in an explicit manner when a hydropower project is ex ante assessed in economic terms. Conventionally, the 90 percent dependable hydrological level is used as a forecast of water flow in the future. However, it is shown that the hydrological forecasts calibrated from the empirical VAR models are very different from the conventional projections.

The climate change impacts in principal differ from location to location; but as far as the selected three projects are concerned, hydrological discharges tend to increase with rainfall and decrease with temperature. It is also shown that the rainy season would likely have higher water levels, but in the lean season water resources would become even more limited.

The resultant This article is about the resultant of polynomials. For the result of adding two or more vectors, see Parallelogram rule. For the technique in organ building, see Resultant (organ).

In mathematics, the resultant of two monic polynomials
 effect on project viability may be modest at best, largely because of the calculus calculus, branch of mathematics that studies continuously changing quantities. The calculus is characterized by the use of infinite processes, involving passage to a limit—the notion of tending toward, or approaching, an ultimate value.  nature of discounting possible costs and benefits in the future. However, when comparing the three cases, it is indicated that having larger installed capacity might be useful to exploit increased hydrological resources for power generation. Also, hydropower stations with some storage capacities may have the advantage of accommodating increased seasonality in hydrological series. However, these may not be able to be overgeneralized, because the paper investigates only three cases. More case studies are necessary for drawing general implications, such as hydropower design alternatives.

Nonetheless, these mitigation measures against uncertain climate changes must have a cost implication in economic and social terms. Hence, a broad and consistent assessment will be needed at the project preparation stage.


Akaike, Hirotugu. 1973. Information theory and an extension of the maximum likelihood principle, in Second International Symposium on Information Theory, eds. by B. N. Petrov and F. Csaki, pp. 267-281, Academiai Kiado, Budapest.

Bergstrom, Sten, Bengt Carlsson, Marie Gardelin, Goran Lindestrom, Anna Pettersson, Markku Rummukainen. 2001. Climate change impacts on runoff in Sweden--assessments by global climate models, dynamical downscaling Global climate models (GCMs) are run at coarse spatial resolution (typically of the order 50,000 km²) and are unable to resolve important sub-grid scale features such as clouds and topography. As a result GCMs can’t be used for local impact studies.  and hydrological modeling. Climate Research, Vol. 16, pp. 101-112.

Bhunya, P.K., R. Berndtsson, C.S.P. Ojha, S.K. Mishra. 2007. Suitability of gamma, chis-quare, weibull, and beta distributions as synthetic unit hydrographs. Journal of Hydrology, Vol. 334, pp. 28-38.

Goswami, B.N., V. Venugopal, D. Sengupta, M.S. Madhusoodanan, P.K. Xavier. 2006. Increasing trend of extreme rain events over India in a warming environment. Science, Vol. 314, pp. 1442-1445.

Kay, Alison, Richard Jones, Nicholas Reynard. 2006. RCM RCM Reliability-Centered Maintenance
RCM Royal College of Music
RCM Royal Conservatory of Music
RCM Royal Canadian Mint
RCM Reliability Centered Maintenance
RCM Revenue Cycle Management
RCM Regional Climate Model
RCM Ring-Closing Metathesis
 rainfall for UK flood frequency estimation
This article is about the technique in signal processing. The term "frequency estimation" can also refer to probability estimation.

Frequency estimation
. II. Climate change results. Journal of Hydrology, Vol. 318, pp. 163-172.

Khaliq, M.N., T.B.M.J. Ouarda, J.-C. Ondo, P. Gachon, B. Bobee. 2006. Frequency analysis of a sequence of dependent and/or non-stationary hydro-meteorological observations: a review. Journal of Hydrology, Vol. 329, pp. 534-442.

Koutsoyiannis, Demetris. 2006. Nonstationarity versus scaling in hydrology. Journal of Hydrology, Vol. 324, pp. 239-254.

Lettenmaier, Dennis, Eric Wood. 1993. Hydrologic forecasting, in Handbook of Hydrology, eds. by David Maidment, pp. 19.1-19.72, McGraw-Hill, Inc., New York New York, state, United States
New York, Middle Atlantic state of the United States. It is bordered by Vermont, Massachusetts, Connecticut, and the Atlantic Ocean (E), New Jersey and Pennsylvania (S), Lakes Erie and Ontario and the Canadian province of

Malamud, Bruce, Donald Turcotte. 2006. The applicability of power-law frequency statistics to floods. Journal of Hydrology, Vol. 322, pp. 168-180.

O'Connell, Daniel. 2005. Nonparametric Bayesian flood frequency estimation. Journal of Hydrology, Vol. 313, pp. 79-96.

Salas, Jose. 1993. Analysis and modeling of hydrologic time series, in Handbook of Hydrology, eds. by David Maidment, pp. 19.1-19.72, McGraw-Hill, Inc., New York.

Samaniego, Luis, Andras Bardossy. 2007. Relating macroclimatic circulation patterns with characteristics of floods and droughts at the mesoscale. Journal of Hydrology, Vol. 335, pp. 109-123.

Shalizi, Zmarak. 2007. Energy and emissions: local and global effects of the rise of China and India. Policy Research Working Paper No. 4209, the World Bank.

Stedinger, Jery, Richard Vogel, Efi Foufoula0Georgiou. 1993. Frequency analysis of extreme events, in Handbook of Hydrology, eds. by David Maidment, pp. 18.1-18.66, McGraw-Hill, Inc., New York.

World Bank. 2007. Addressing climate change-driven increased hydrological variability in environmental assessments for hydropower projects--a scoping study. Mimeograph.

Wong, Heung, Wai-cheung Ip, Riquan Zhang, Jun Xia. 2007. Hon-parametric time series models for hydrologic forecasting. Journal of Hydrology, Vol. 332, pp. 337-347.

(1) The project economic analysis usually accounts for economic costs and benefits in 30 years.

(2) In these time series borrowed from the NOAA database, some observations are missing in the 1990s.

(3) There might be a much longer term trend in climate time series, which is not taken into account in the current analysis. Goswami et al. (2006) show that India's precipitation time series exhibits a long-term trend over the last 50 years.

(4) The Mukteshwar observatory (lat. 29.5[degrees] north; long. 79.7[degrees] east) provides data to the Vishnugad Pipalkoti case; and data from Colombo (lat. 6.9[degrees] north; long. 79.9[degrees] east) are borrowed.

(5) The latitude and longitude latitude and longitude

Coordinate system by which the position or location of any place on the Earth's surface can be determined and described. Latitude is a measurement of location north or south of the Equator.
 of the specified point are 12[degrees] north and 107[degrees] east, respectively.

(6) The NOAA database provides precipitation time series per day, which is thus converted to monthly series in the Thac Mo case.

(7) These figures capture only the deterministic trend component. Total increment of temperature will be determined by the other deterministic and stochastic dynamics in the system.

(8) It is a separate question whether the past hydrological time series contain useful information.

(9) The dynamic forecasts tend to have much large standard errors, due to the nature of the estimation method. Since we are computing computing - computer  the 240-period-ahead projections, the standard errors are amplified quickly.

(10) Again, the baseline is a conservative assumption, which is the 10th percentile estimate. On the other hand, our dynamic estimate is, roughly speaking, calculated on a mean basis, as usual.

(11) For simplicity, it is also assumed that the total utilization rate of installed capacity is 95 percent. In addition, the combined turbine and generator efficiency is commonly assumed 93 percent.

(12) Extreme events induced by increased variability and seasonality are a different issue and beyond the scope of this sturdy. This paper is analyzing hydrological series merely on the monthly mean level.

(13) Note that the storage capacity attached to the Upper Kotmale power station is too small to influence the current discussion.

(14) The assumed operating rule of the storage capacity is this: extra water is stored in a reservoir whenever the available water flow exceeds the hydroplant deign deign  
v. deigned, deign·ing, deigns

To think it appropriate to one's dignity; condescend: wouldn't deign to greet the servant who opened the door.
 discharge. When the available water is insufficient, the stored water is used for generation as long as there is water.

(15) The development of international support mechanisms for climate change adaptation seems to be lagging Lagging

Strategy used by a firm to stall payments, normally in response to exchange rate projections.
 behind climate change mitigation. Adaptation measures are mostly considered private goods, while mitigation efforts can be rewarded through the growing carbon market, due to their perceived positive global externalities externalities

side-effects, either harmful or beneficial, borne by those not directly involved in the production of a commodity.
. In the case of hydropower, however, there are synergies between the adaptation and mitigation agendas: additional installed generation or reservoir capacities could help to adapt to expected changes in river flows and also result in increased production of electricity with low/zero carbon emissions, replacing carbon-intensive power generation.

(16) Technically, it may be somewhat meaningful to predict the extreme river flows based on our estimated standard error. For instance, the predicted flows at the 90 percent dependable level--meaning an upper bound of the less significant interval--could be interpreted as a very unlikely flood event in a statistical sense.

(17) If all time series in the system contain a unit root, the vector error-correction model is more appropriate.

(18) The Himalayan glaciers are currently retreating at a speed of 10-15 meters a year (WWF See Windows Workflow Foundation.  Nepal Program, 2005).

The World Bank Sustainable Development Sustainable development is a socio-ecological process characterized by the fulfilment of human needs while maintaining the quality of the natural environment indefinitely. The linkage between environment and development was globally recognized in 1980, when the International Union  Network Finance, Economics, and Urban Development Department September 2007

This paper--a product of the Finance, Economics, and Urban Development Department--is part of a larger effort in the department to examine infrastructure development and climate changes. Policy Research Working Papers working papers
Legal documents certifying the right to employment of a minor or alien.

Noun 1. working papers
 are also posted on the Web at The author may be contacted at

Estimating Global Climate Change Impacts on Hydropower Projects: Applications in India, Sri Lanka and Vietnam

Atsushi IIMI IIMI International Irrigation Management Institute
IIMI Indian Institute of Management, Indore (India) 

Finance, Economics and Urban Finance (FEU)

The World Bank

1818 H Street N.W. Washington D.C. 20433

Tel: 202-473-4698 Fax: 202-522-3481


([para]) I am most grateful to Kenneth Chomitz, Antonio Estache, Michael Haney, Charles Kenney, Laszlo Lovei, Alessandro Palmieri and Winston Yu for their insightful suggestions on an earlier version of this paper. I also aknowledge JBIC JBIC Japan Bank for International Cooperation
JBIC Japan Biological Informatics Consortium
 colleagues, Keiko Kuroda, Keiju Mitsuhashi, Yasuhisa Ojima and Kazuko Tatsumi.
Table 1. Summary of Project Description

                             Vishnugad               Upper
                             Pipalkoti               Kotmale

Country                      India                   Sri Lanka

Water source                 Alaknanda               Kotmale
                             River                   River
Catchment area               2,700                   ...

Project location             Lat. 30.3[degrees] N;   Lat. 7.0[degrees]
                             Lon. 79.3[degrees] E    N; Lon.
                                                     80.5[degrees] E

New/extension project        New                     New

Installed capacity (MW)      440                     150

Annual generated energy      1,838                   528
(GWh) (1/)

Facility utilization         100                     40
rate (%)

Availability of installed    95                      95
capacity (%)

90% firm monthly
generation capacity (MW)

  Average                    212                     46

  Maximum                    440                     107

  Minimum                    70                      6

Average available water      115                     10
flows ([m.sup.3]/sec)

Net head (m)                 205                     473

Intended power generation    Base load (wet)         Peak load
operation                    Peak load (dry)

Reservoir storage capacity   3,630                   800
([10.sup.3] [m.sup.3])

Relative to the average      4 hours                 1 day
water flow

Hydrological data            Jun 1974 to             Oct 1951 to
availability                 May 2004                Sep 1998

                             Thac Mo

Country                      Vietnam

Water source                 Thac Mo

Catchment area               2,200

Project location             Lat. 12[degrees] N;
                             Lon. 107[degrees] E

New/extension project        Extension

Installed capacity (MW)      75

Annual generated energy      52
(GWh) (1/)

Facility utilization         39
rate (%)

Availability of installed    92
capacity (%)

90% firm monthly
generation capacity (MW)

  Average                    48

  Maximum                    75

  Minimum                    10

Average available water      83
flows ([m.sup.3]/sec)

Net head (m)                 90

Intended power generation    Peak load

Reservoir storage capacity   1,250,000
([10.sup.3] [m.sup.3])

Relative to the average      6 months
water flow

Hydrological data            Jan 1976 to
availability                 Dec 2001

Sources: Feasibility studies.

(1/) The annual energy estimates are the original figures in the
individual feasibility studies, which account for not only water
resources availability but also other operational factors in the grid,
including dams and other types of power plants.

Table 2. IPCC Climate Projections around Project Sites

            Vishinugad Pipalkoti (India)

      Temperature ([degrees]C)   (mm/m month)

      Avg.               Proj.   Avg.      Proj.
      1961-90            2020s   1961-90   2020s
Jan    6.3                7.3     47        34
Feb    7.2                8.0     55        47
Mar   11.0               12.2     58        39
Apr   15.5               17.2     35         6
May   17.8               19.7     66        52
Jun   18.6               20.0    138       135
Jul   17.4               18.2    327       340
Aug   17.0               17.4    293       332
Sep   16.2               16.3    201       198
Oct   14.3               15.1     43        56
Nov   11.1               11.5      7         0
Dec    8.4                9.6     24        27

              Upper Kotmale (Sri Lanka)

      Temperature ([degrees]C)   (mm/m month)

      Avg.               Proj.   Avg.      Proj.
      1961-90            2020s   1961-90   2020s

Jan   26.6               27.3     63        97
Feb   26.9               27.6     71        68
Mar   27.7               28.3    129       137
Apr   28.2               28.8    255       252
May   28.3               28.9    401       335
Jun   27.9               28.7    179       187
Jul   27.6               28.4    130       138
Aug   27.5               28.3     92       111
Sep   27.5               28.3    241       289
Oct   27.0               27.6    382       379
Nov   26.7               27.4    308       285
Dec   26.6               27.2    170       156

                  Thac Mo (Vietnam)

      Temperature ([degrees]C)   (mm/m month)

      Avg.               Proj.   Avg.      Proj.
      1961-90            2020s   1961-90   2020s

Jan   21.5               21.9     44        51
Feb   22.5               22.7     26        38
Mar   20.0               20.8     34        73
Apr   25.5               26.1     48        61
May   26.2               26.3    122       135
Jun   26.5               26.4    154       167
Jul   26.6               27.0    170       135
Aug   26.8               27.2    160       168
Sep   25.7               26.3    311       313
Oct   25.1               25.7    363       397
Nov   23.7               24.3    312       309
Dec   22.2               23.1    110       113

Sources: IPCC DDC database and one of the SRES scenarios, CSIRO/A1a;
and NOAA GHCN Monthly database version 2.

Note that the estimated incremental changes are given by the IPCC
model. The historical series are based on NOAA database for
VPHEP and UKHPP. The Thac Mo case relies on IPCC DDC database for
historical data as well.

Table 3. Unit Root Test

                 Vishnugad     Upper
                 Pipalkoti     Kotmale       Thac Mo
                 (India)       (Sri Lanka)   (Vietnam)

lnHYDRO           -4.324 ***   -4.166 ***    -1.827

lnTEMP            -5.432 ***   -3.656 ***    -3.286 **

lnPREC            -6.469 ***   -4.619 ***    -3.645 ***

[DELTA]lnHYDRO    -6.128 ***   -9.805 ***    -6.400 ***

[DELTA]lnTEMP    -15.860 ***   -7.667 ***    -6.780 ***

[DELTA]lnPREC    -16.656 ***   -8.573 ***    -8.592 ***

Note: *** 1 % significance level, ** 5 % significance level,
and * 10% significance level.

Table 4. Simple Correlation

                    Vishnugad   Upper
                    Pipalkoti   Kotmale       Thac Mo
                    (India)     (Sri Lanka)   (Vietnam)

(lnHYDRO, lnTEMP)   0.774       -0.360        0.465

(lnHYDRO, lnPREC)   0.271        0.285        0.849

(lnTEMP, lnPREC)    0.117        0.030        0.551

No of obs.          295         496           240

Table 5. Lag Selection Criteria

                Vishnugad Pipalkoti (India)

              W/o trend                With trend
      Log          AIC          Log          AIC
Lag   likelihood   statistics   likelihood   statistics

0     -349.9        3.820       -339.0        3.743
1     -215.8        2.582       -209.1        2.545
2     -203.2        2.546 **    -197.5        2.520 **
3     -197.5        2.579       -192.3        2.558
4     -191.3        2.607       -185.9        2.583

                 Upper Kotmale (Sri Lanka)

              W/o trend                With trend
      Log          AIC          Log          AIC
Lag   likelihood   statistics   likelihood   statistics

0      110.1       -0.370        184.0       -0.723
1      256.8       -1.056        279.7       -1.155
2      286.9       -1.162        299.4       -1.209
3      303.5       -1.199 **     311.2       -1.223 **
4      307.1       -1.172        315.0       -1.197

                     Thac Mo (Vietnam)

              W/o trend                With trend
      Log          AIC          Log          AIC
Lag   likelihood   statistics   likelihood   statistics

0      673.8       -5.405        696.3       -5.570
1      737.8       -5.871 **     747.0       -5.924 **
2      741.7       -5.828        749.5       -5.869
3      751.2       -5.832        756.6       -5.852
4      756.3       -5.799        761.0       -5.814

Note: *** 1 % significance level, ** 5 % significance level,
and * 10% significance level.

Table 6. Estimated Hydro Equation: Vishnugad Pipalkoti, India

                                         W/o trend

                                 Unrestricted   Restricted

lnHYDRO (t-1)                      1.072 ***      1.093 **
                                  (0.064)        (0.054)

lnHYDRO (t-2)                     -0.209 ***     -0.251 **
                                  (0.068)        (0.059)

lnTEMP (t-1)                      -0.817 **      -0.671 **
                                  (0.374)        (0.238)

lnTEMP (t-2)                       0.236

lnPREC(t-1)                       -0.003

lnPREC(t-2)                        0.008 **       0.007 *
                                  (0.003)        (0.003)

[c.sub.February]                   0.018

[c.sub.March]                      0.233 **       0.181 **
                                  (0.093)        (0.069)

[c.sub.April]                      0.656 ***      0.590 **
                                  (0.104)        (0.078)

[c.sub.May]                        1.168 ***      1.108 **
                                  (0.130)        (0.100)

[c.sub.June]                       0.944 ***      0.908 **
                                  (0.162)        (0.110)

[c.sub.July]                       0.810 ***      0.811 **
                                  (0.170)        (0.091)

[c.sub.August]                     0.419 **       0.442 **
                                  (0.168)        (0.076)

[c.sub.September]                  0.077

[c.sub.October]                   -0.228

[c.sub.November]                   0.021

[c.sub.December]                  -0.079


constant                           2.758 *        3.230 **
                                  (1.674)        (0.807)

Obs.                             238            238

R-squared                          0.951          0.946

Normality test

  Skewness statistics              6.764 **       4.900 *

Eigenvalue stability condition     6.764

  Max modulus                      0.739          0.765

Autocorrelation test

  LM test statistics              18.695 **      10.293

                                         W/o trend

                                 Unrestricted   Restricted

lnHYDRO (t-1)                      1.066 **       1.094 ***
                                  (0.065)        (0.054)

lnHYDRO (t-2)                     -0.212 **      -0.249 ***
                                  (0.068)        (0.059)

lnTEMP (t-1)                      -0.849 **      -0.681 ***
                                  (0.375)        (0.238)

lnTEMP (t-2)                       0.204

lnPREC(t-1)                       -0.003

lnPREC(t-2)                        0.009 **       0.007 **
                                  (0.003)        (0.003)

[c.sub.February]                   0.008

[c.sub.March]                      0.218 ***      0.183 ***
                                  (0.095)        (0.069)

[c.sub.April]                      0.645 ***      0.594 ***
                                  (0.104)        (0.078)

[c.sub.May]                        1.168 ***      1.111 ***
                                  (0.130)        (0.100)

[c.sub.June]                       0.960 ***      0.911 ***
                                  (0.162)        (0.111)

[c.sub.July]                       0.834 ***      0.812 ***
                                  (0.172)        (0.091)

[c.sub.August]                     0.446 ***      0.441 ***
                                  (0.171)        (0.076)

[c.sub.September]                  0.104

[c.sub.October]                   -0.205

[c.sub.November]                   0.036

[c.sub.December]                  -0.069

t                                  0.0001

constant                           3.020 *        3.254 ***
                                  (1.697)        (0.807)

Obs.                             238            238

R-squared                          0.951          0.946

Normality test

  Skewness statistics              4.313 **       5.053 **

Eigenvalue stability condition

  Max modulus                      0.752          0.770

Autocorrelation test

  LM test statistics              20.031 **      10.532

Note that the dependent variable is lnHYDRO. The standard errors are
shown in parentheses. *** 1 % significance level, ** 5 % significance
level, and * 10% significance level.

Table 7. Estimated Hydro Equation: Upper Kotmale, Sri Lanka

                                        W/o trend

                                 Unrestricted   Restricted

lnHYDRO (t-1)                      0.418 **       0.412 ***
                                  (0.052)        (0.047)

lnHYDRO (t-2)                      0.092 *        0.156 ***
                                  (0.057)        (0.045)

lnHYDRO(t-3)                      -0.034

lnTEMP (t-1)                      -3.399         -5.479 ***
                                  (2.554)        (2.004)

lnTEMP (t-2)                      -3.094

lnTEMP (t-3)                      (2.033)

lnPREC(t-1)                        0.016 *        0.016 *
                                  (0.009)        (0.009)

lnPREC(t-2)                        0.013

lnPREC(t-3)                        0.008

[c.sub.February]                  -0.250 **      -0.207 ***
                                  (0.105)        (0.086)

[c.sub.March]                     -0.054

[c.sub.April]                      0.284 **       0.374 ***
                                  (0.133)        (0.096)

[c.sub.May]                        0.750 ***      0.841 ***
                                  (0.148)        (0.110)

[c.sub.June]                       0.930 ***      0.981 ***
                                  (0.141)        (0.118)

[c.sub.July]                       0.962 ***      0.941 ***

                                  (0.132)        (0.111)
[c.sub.August]                     0.867 ***      0.805 ***
                                  (0.125)        (0.105)

[c.sub.September]                  0.679 ***      0.626 ***
                                  (0.120)        (0.109)

[c.sub.October]                    0.911 ***      0.841 ***
                                  (0.117)        (0.104)

[c.sub.November]                   0.724 ***      0.654 ***
                                  (0.115)        (0.101)

[c.sub.December]                   0.350 ***      0.314 ***
                                  (0.107)        (0.096)


constant                          38.641 ***     25.022 ***
                                 (11.238)        (8.849)

Obs.                             416            416

R-squared                          0.665          0.660

Normality test

  Skewness statistics              3.350 *        1.770

Eigenvalue stability condition     3.350

  Max modulus                      0.886          0.880

Autocorrelation test

  LM test statistics              18.901 **       9.138

                                        With trend

                                 Unrestricted   Restricted

lnHYDRO (t-1)                      0.421 ***      0.407 ***
                                  (0.052)        (0.047)

lnHYDRO (t-2)                      0.093 *        0.148 ***
                                  (0.057)        (0.045)

lnHYDRO(t-3)                      -0.040

lnTEMP (t-1)                      -2.612         -6.180 ***
                                  (2.645)        (2.015)

lnTEMP (t-2)                      -2.407

lnTEMP (t-3)                      -1.301

lnPREC(t-1)                        0.016 *        0.011
                                  (0.009)        (0.008)

lnPREC(t-2)                        0.013

lnPREC(t-3)                        0.008

[c.sub.February]                  -0.244 **      -0.213 **
                                  (0.105)        (0.086)

[c.sub.March]                     -0.052

[c.sub.April]                      0.268 **       0.382 ***
                                  (0.134)        (0.096)

[c.sub.May]                        0.706          0.861 ***
                                  (0.153) ***    (0.110)

[c.sub.June]                       0.861          1.008 ***
                                  (0.154) ***    (0.118)

[c.sub.July]                       0.890          0.965 ***
                                  (0.147) ***    (0.111)

[c.sub.August]                     0.807          0.826 ***
                                  (0.136) ***    (0.105)

[c.sub.September]                  0.632          0.649 ***
                                  (0.127) ***    (0.109)

[c.sub.October]                    0.873          0.865 ***
                                  (0.121) **     (0.104)

[c.sub.November]                   0.692          0.673 ***
                                  (0.118) ***    (0.101)

[c.sub.December]                   0.334          0.328 ***
                                  (0.108) ***    (0.095)

t                                 -0.0002

constant                          29.019 **      28.155 ***
                                 (14.121)        (8.897)

Obs.                             416            416

R-squared                          0.666          0.660

Normality test

  Skewness statistics              3.249 *        1.795

Eigenvalue stability condition

  Max modulus                      0.778          0.764

Autocorrelation test

  LM test statistics              14.855 *       11.330

Note that the dependent variable is lnHYDRO. The standard errors are
shown in parentheses. *** 1 % significance level, ** 5 % significance
level, and * 10% significance level.

Table 8. Estimated Hydro Equation: Thac Mo, Vietnam

                                        W/o trend

                                 Unrestricted   Restricted

lnHYDRO (t-1)                      0.532 **       0.657 **
                                  (0.056)        (0.030)

lnTEMP (t-1)                       2.145

lnPREC (t-1)                       0.097

[c.sub.February]                  -0.158

[c.sub.March]                     -0.124

[c.sub.April]                      0.057

[c.sub.May]                        0.713 ***      1.041 ***
                                  (0.220)        (0.093)

[c.sub.June]                       1.231 ***      1.581 ***
                                  (0.178)        (0.086)

[c.sub.July]                       1.517 ***      1.762 ***
                                  (0.189)        (0.088)

[c.sub.August]                     1.820 ***      1.958 ***
                                  (0.206)        (0.099)

[c.sub.September]                  1.525 ***      1.567 ***
                                  (0.231)        (0.113)

[c.sub.October]                    1.237 ***      1.271 ***
                                  (0.242)        (0.116)

[c.sub.November]                   0.478 **       0.518 ***
                                  (0.238)        (0.112)

[c.sub.December]                   0.095


constant                          -8.273          0.779 ***
                                  (6.550)        (0.122)

Obs.                             239            239

R-squared                          0.947          0.943

Normality test

  Skewness statistics             31.928 ***     21.081 ***

Eigenvalue stability condition

  Max modulus                      0.536          0.657

Autocorrelation test

  LM test statistics               7.404         16.901 *

                                       With trend

                                 Unrestricted   Restricted

lnHYDRO (t-1)                      0.481 ***      0.642 ***
                                  (0.060)        (0.030)

lnTEMP (t-1)                       0.849

lnPREC (t-1)                       0.111

[c.sub.February]                  -0.207

[c.sub.March]                     -0.172

[c.sub.April]                     -0.096

[c.sub.May]                        0.728 ***      1.022 ***
                                  (0.218)        (0.093)

[c.sub.June]                       1.300 ***      1.575 ***
                                  (0.179)        (0.085)

[c.sub.July]                       1.648 ***      1.773 ***
                                  (0.196)        (0.088)

[c.sub.August]                     1.996 ***      1.981 ***
                                  (0.217)        (0.099)

[c.sub.September]                  1.747 ***      1.602 ***
                                  (0.248)        (0.113)

[c.sub.October]                    1.424 ***      1.307 ***
                                  (0.253)        (0.116)

[c.sub.November]                   0.630 ***      0.551 ***
                                  (0.245)        (0.112)

[c.sub.December]                   0.163

t                                  0.0008 **      0.0007 **
                                  (0.0003)       (0.0003)

constant                          -2.804          0.634 ***
                                  (6.907)        (0.146)

Obs.                             239            239

R-squared                          0.948          0.944

Normality test

  Skewness statistics             28.797 ***     19.477 ***

Eigenvalue stability condition

  Max modulus                      0.471          0.642

Autocorrelation test

  LM test statistics               4.209         12.995

Note that the dependent variable is lnHYDRO. The standard errors are
shown in parentheses. *** 1 % significance level, ** 5 % significance
level, and l0% significance level.

Table 9. Time Trend Coefficients from Unrestricted Models

                      Vishnugad    Upper
                      Pipalkoti    Kotmale         Thac mo
                     (India)      (Sri Lanka)     (Vietnam)

lnHYDRO equation:
  t                  0.00011      -0.00021         0.00079 **
                    (0.00013)     (0.00018)       (0.00035)

lnTEMP equation:
  t                  0.000050 *    0.000015 ***    0.000042 ***
                    (0.000023)    (0.000004)      (0.000014)

lnPREC equation:
  t                 -0.00517 *    -0.00178 *      -0.00016
                    (0.00272)     (0.00098)       (0.00015)

The standard errors are shown in parentheses. *** 1 % significance
level, ** 5 % significance level, and * 10% significance level.

Table 10. Causality Test

                               Vishnugad    Upper
                               Pipalkoti    Kotmale       Thac Mo
                               (India)      (Sri Lanka)   (Vietnam)
                               Chi2         Chi2          Chi2
Null hypothesis                statistics   statistics    statistics

lnTEMP [right arrow] lnHYDRO   4.789 *      11.508 ***     1.990

lnPREC [right arrow] lnHYDRO   6.935 *       5.788         0.450

lnHYDRO [right arrow] lnTEMP   4.046         6.485 *       0.003

lnPREC [right arrow] lnTEMP    4.077         3.816         0.321

lnHYDRO [right arrow] lnPREC   0.427         1.364         1.928

lnTEMP [right arrow] lnPREC    2.159         1.554        12.665 ***

*** 1 % significance level, ** 5 % significance level, and * 10%
significance level.

Table 11. Impact of Changes in Hydrology on Electricity Generation

                          Vishnugad Pipalkoti   Upper Kotmale
                          (India)               (Sri Lanka)

                          Baseline   Dynamic    Baseline   Dynamic
                                     forecast              forecast
                                     (2025)                (2025)

Annual energy (GWh)       1,768      1,898      357        522

Storage capacity          No         No         No         No

Memorandum items:

Installed capacity (MW)   440        440        150        150

Monthly available
capacity (MW)

  Average                 212        280        46         67

  Maximum                 440        440        107        94

  Minimum                 70         63         6          24

Average available water   115        156        10         14
flows ([m.sup.3]/sec)

Net head (m)              205        205        473        473

Intended power            Base       Base       Base       Base
generation operations

                                        Thac Mo

                          Baseline   Dynamic    Dynamic
                                     forecast   forecast
                                     (2025)     (2025)

Annual energy (GWh)       383        331        524

Storage capacity          No         No         Yes

Memorandum items:

Installed capacity (MW)   75         75         75

Monthly available
capacity (MW)

  Average                 48         44         62

  Maximum                 75         75         75

  Minimum                 10         5          5

Average available water   83         88         81
flows ([m.sup.3]/sec)

Net head (m)              90         90         90

Intended power            Base       Base       Base
generation operations

Source: Author's estimates.

Note: The baseline scenario assumes hydrological series in a 90
percent dependable year.

Table 12. Impact of Changes in Hydrology on Internal Rate of Return

                                  Original project cost

                                         Dynamic forecasts (2025)

                              Baseline   Without   With current
                                         storage   storage

Vishnugad Pipalkoti (India)   15.8%      16.1%     16.1%

Upper Kotmale (Sri Lanka)      4.7%       6.4%      6.4%

Thac Mo (Vietnam)             29.0%      28.8%     29.6%

                              50% additional project cost
                              for a 6-month storage capacity

                              Baseline   forecasts (2025)

Vishnugad Pipalkoti (India)   10.9%      11.9%

Upper Kotmale (Sri Lanka)      1.2%       3.2%

Thac Mo (Vietnam)             21.9%      22.6%

Source: Author's estimates.

Note: The baseline scenario assumes hydrological series in a 90
percent dependable year without major storage capacity.
COPYRIGHT 2007 The World Bank
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2007 Gale, Cengage Learning. All rights reserved.

 Reader Opinion




Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Policy Research Working Paper
Author:Iimi, Atsushi
Publication:Estimating Global Climate Change Impacts on Hydropower Projects: Applications in India, Sri Lanka an
Date:Sep 1, 2007

Terms of use | Copyright © 2014 Farlex, Inc. | Feedback | For webmasters