COMPARISON OF REANALYSIS DATA AND METEOROLOGICAL STATION DATA USING RAINFALL DATA FOR DROUGHT ASSESSMENT IN MAHALAPYE AREA BY OGOROGILE MARUMO

COMPARISON OF REANALYSIS DATA AND METEOROLOGICAL STATION
DATA USING RAINFALL DATA FOR DROUGHT ASSESSMENT IN MAHALAPYE
AREA
BY

OGOROGILE MARUMO
(ID: 14001849)

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

A PROJECT SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS
FOR THE DEGREE OF BACHELOR OF SCIENCE (BSc
ENVIRONMENTALSCIENCES) IN THE
DEPARTMENT OF EARTH AND ENVIRONMENTAL SCIENCES
COLLEGE OF SCIENCE

AUGUST 2018

1

DECLARATION
I do hereby declare that this research project “Comparison of reanalysis data and
meteorological station data using rainfall data for drought assessment in Mahalapye” herein
presented for a Bachelor of Degree of Science (BSc) in Environmental Science is a
result of my investigations. To the best of my believe and knowledge, it has not been
presented in its current form for a Degree elsewhere. All reference to other
researchers’ works as sources of information are duly acknowledged

…………………………
Ogorogile Marumo
Student
Date
This declaration is affirmed by
…………………………..
Dr SITHABILE TIRIVAROMBO
(PROJECT SUPERVISOR)
DATE:

2

ACKNOWLEDGEMENTS

I would like to thank my supervisor for helping me out with the project. Also I would
like to thank my classmates for supporting me during the course of the project

3

ABSTRACT

To monitor climate variables, studies shows that there are two sources of data that
can be used. These are, ground measurements from weather stations and satellite
observations. The weather station data is very accurate at recording what is happening
at ground level but only in the immediate vicinity of the station. However, weather
stations are expensive and require constant recording of measurements and periodic
maintenance. Also, weather stations are sparsely distributed, especially in developing
countries. Rural areas are poorly covered and require interpolation between stations.
In contrast, satellites see entire landscapes and so they are able to provide precise
measurements at every location. However, satellites cannot make precipitation ground
measurements as accurately as weather stations (Mendelsoh et al 2007). Due to
limitations of the station observed data, this study aims to explore the use of reanalysis
data in addressing water resources problems. The rainfall data was obtained from the
Mahalapye meteorology station. R is a programming language and environment used
for statistical computing and graphics. The Mahalapye data was sorted into the
variables required by the calculator before putting it into the calculator. Thereafter it
was put into the calculator and the calculator produced the output in the form of
spreadsheet data and graphs for both the SPI and SPEI. The data outputs for all the
drought indicators (SPI and SPEI) shows that the year 1992 was hit by an extreme
drought represented by the values ranging from -1.5 to -2.4 of the SPI index
description. The results shows that there is a small difference between the reanalysis
data and a weather station observed data and the difference is brought up by the
difference in areal cover of the two data sources/models

4

Table of Contents
DECLARATION ………………………………………………………………………………………………………………….. 1
ACKNOWLEDGEMENTS …………………………………………………………………………………………………… 2
ABSTRACT…………………………………………………………………………………………………………………………. 3
LIST OF TABLES ……………………………………………………………………………………………………………….. 6
Table 1: shows the annual average temperature recorded at Mahalapye weather station …… 6
Table 2: shows the monthly precipitation and annual average precipitation recorded at
Mahalapye meteorological station ……………………………………………………………………………………….. 6
Table 3: SPI indices from station data …………………………………………………………………………………. 6
Table 4: SPEI indices for station data ………………………………………………………………………………….. 6
Table 5: SPI indices for reanalysis data ………………………………………………………………………………. 6
Table 6: SPEI indices for reanalysis data …………………………………………………………………………….. 6
Table 7: the SPI index values ………………………………………………………………………………………………. 6
TABLE 8. Interpretation of SPI values …………………………………………………………………………………. 6
TABLE 9. Interpretation of SPEI values ……………………………………………………………………………….. 6
LIST OF FIGURES ……………………………………………………………………………………………………………… 7
Figure 1: example of an automatic weather station ……………………………………………………………… 7
CHAPTER ONE ………………………………………………………………………………………………………………….. 8
INTRODUCTION …………………………………………………………………………………………………………………. 8
BACKGROUND INFORMATION ……………………………………………………………………………………… 8
PROBLEM STATEMENT …………………………………………………………………………………………………. 9
MAIN OBJECTIVE ……………………………………………………………………………………………………… 10
SPECIFIC OBJECTIVES ……………………………………………………………………………………………. 10
RESEARCH QUESTIONS ………………………………………………………………………………………….. 10
LITERATURE REVIEW ………………………………………………………………………………………………….. 11
WEATHER STATION OBSERVED DATA ………………………………………………………………….. 14
REANALYSIS DATA …………………………………………………………………………………………………… 16
THE WATER CYCLE INTEGRATOR (WCI) ……………………………………………………………….. 20
CHAPTER TWO ………………………………………………………………………………………………………………… 22
MATERIALS AND METHODS ………………………………………………………………………………………….. 22
THE STUDY AREA ………………………………………………………………………………………………………… 22
DATA AND METHODS ………………………………………………………………………………………………….. 23
CHAPTER THREE …………………………………………………………………………………………………………….. 25
RESULTS ………………………………………………………………………………………………………………………….. 25
CHAPTER FOUR ………………………………………………………………………………………………………………. 35
DISCUSSION …………………………………………………………………………………………………………………….. 35

5

CHAPTER FIVE ………………………………………………………………………………………………………………… 37
CONCLUSION …………………………………………………………………………………………………………………… 37
REFERENCES ………………………………………………………………………………………………………………….. 38

6

LIST OF TABLES

Table 3.1: shows the annual average temperature recorded at Mahalapye weather
station
Table 3.2: shows the monthly precipitation and annual average precipitation recorded
at Mahalapye meteorological station
Table 3.3: SPI indices from station data
Table 3.4: SPEI indices for station data
Table 3.5: SPI indices for reanalysis data
Table 3.6: SPEI indices for reanalysis data
Table 3.7: the SPI index values
Table 3.8. Interpretation of SPI values
Table 3.9. Interpretation of SPEI values

7

LIST OF FIGURES

Figure 1.1: example of an automatic weather station

Figure 2.1: shows the study area of Mahalapye

8

CHAPTER ONE
INTRODUCTION

Monitoring changes in the Earth’s climate is based on long time series of atmospheric
and ocean observations. Included among these are long time instrumental
measurements of surface temperature and precipitation and records of daily or
monthly data which are important in understanding changes in the frequency and
severity of extreme events such as heavy floods, drought and heat waves.(Wang et al
2008)
A climate monitoring system integrates satellite observations, ground-based data and
forecast models to monitor and forecast changes in the weather and climate. A
historical record of spot measurements is built up over time, which provides the data
to enable statistical analysis and the identification of mean values, trends and
variations (Hylke et al 2017). The data can be stored in a library of archives so that it
can be retrieved when needed to assess past climate and present climate. The more
the information is available, the more accurately future conditions can be assessed, at
the local, regional, national and global level.
BACKGROUND INFORMATION

Drought is a natural hazard that results from low levels of precipitation than what is
considered normal. When this phenomena extends over a longer period of time,
rainfall is insufficient to meet the demands of human activities and the environment
Through satellite–aerial–ground observation experiments, researchers have acquired
large amounts of data, including multi-platform, multi-band, and multi-scale data
(Mendelsoh et al 2007).

9

Based on these data, researchers have explored new technologies, and methods for
climate change studies, developed assimilation models using multi-source
heterogeneous spatial data. Other scientific data obtained can include fundamental
geographic information data, ground-based and in-situ observation data, and data
derived from Earth system and data assimilation models. Station-based observations
are collected from instruments sited at locations on every continent. They include
temperature, relative humidity, precipitation, wind speed and direction
PROBLEM STATEMENT

To monitor climate variables, studies shows that there are two sources of data that
can be used. These are, ground measurements from weather stations and satellite
observations. The weather station data is very accurate at recording what is happening
at ground level but only in the immediate vicinity of the station. However, weather
stations are expensive and require constant recording of measurements and periodic
maintenance. Also, weather stations are sparsely distributed, especially in developing
countries. Rural areas are poorly covered and require interpolation between stations.
In contrast, satellites see entire landscapes and so they are able to provide precise
measurements at every location. However, satellites cannot make precipitation ground
measurements as accurately as weather stations (Mendelsoh et al 2007). Due to
limitations of the station observed data, this study aims to explore the use of reanalysis
data in addressing water resources problems. In this study reanalysis data is
described as that data which was derived from merging station observed and satellite
derived data sets. The study compares the outputs derived from the reanalysis data,
obtained from the water Cycle Integrator (WCI) with the outputs derived from the data
acquired from Mahalapye meteorology station. The variables which it will be focussing

10

on are the precipitation and precipitation derived drought indicators which are the
Standardized Precipitation Index (SPI) and the Standardized Precipitation Evaporation
Index (SPEI).
MAIN OBJECTIVE

To compare the Reanalysis Project data and the ground meteorology station data for
Mahalapye
SPECIFIC OBJECTIVES

To obtain station observed data from Mahalapye Meteorological Station
To generate SPI and SPEI using data from Mahalapye met station
To obtain reanalysis data from the Water Cycle Integrator (WCI) and use it to generate
SPI and SPEI for Mahalapye
To compare the outputs of reanalysis data and station observed data
RESEARCH QUESTIONS

1. How does the data outputs of reanalysis data compare to the data outputs of
Mahalapye met station data
2. Are there any differences between the outputs of reanalysis data and
Mahalapye meteorological station data

11

LITERATURE REVIEW

Rainfall is a major component of the climate system and plays a key role in the Earth’s
hydrological cycle and energy balance. Thus, accurate measurement of rainfall is vital
to analyze the spatial and temporal patterns of precipitation at various scales.
However, ground rainfall stations in many parts of the world are very sparse and
unevenly distributed (Hylke et al 2017). As a result, analysis using rain gauge
observation is significantly limited to point based particular location. Because of this
scattered distribution of weather stations, the dependability of rain gauge data to
estimate areal rain and spatial distribution of rainfall over large areas is considerably
reduced. To address this limitation, advances in remote sensing science provided an
opportunity to estimate rainfall from satellite observations and are an important source
of rainfall data (Ayehu et al 2018).
Although weather stations give accurate measures of ground conditions, they contain
observations occurring at irregular interval or only in a few places that require
interpolation where observations are missing. On the other hand satellites provide
complete spatial coverage of various parameters over a landscape. For example
satellite temperature measurements slightly outperform the interpolated ground
station data but the precipitation ground measurements generally outperform the
satellite surface wetness index. Satellites provide promising measures of temperature
but ground station data may still be preferred for measuring precipitation in rural
settings. (Mendelsoh et al 2007)
Conventional observations of meteorological variables (atmospheric pressure,
temperature and moisture, precipitation etc.) are collected daily at thousands of
meteorological stations around the world, to be used for weather analysis and

12

forecasting. (Kuleshov 2017). Statistical analysis of the archived data over long-term
period (decades and longer) allows one to derive conclusions about climate based on
instrumental records obtained at meteorological stations. The continuity of records is
crucial for climatological applications, detection of historical trends in meteorological
variables etc. Nonetheless, conventional records are restricted to locations of
meteorological stations. In modern time, data obtained by remote sensing instruments
has revolutionised the science of meteorology and climatology as they provide
potentially global coverage and improve access to areas which have limited number
of meteorological stations (data sparse areas) or not covered by conventional
observations at all. Remote sensing data complement conventional observations and
is widely used in numerical weather prediction, adding value to and improving skill of
weather forecasts.
Satellite-derived rainfall estimates (SREs) are available from thermal infrared radiation
(TIR) and passive microwave (PMW) channels, from geostationary and low Earth-
orbiting satellites, respectively. Techniques for satellite rainfall estimates have
limitation and embedded uncertainties because satellites do not measure rainfall by
itself and should be related to precipitation based on one or multiple surrogate
variables. (Ayehu et al 2018). The uncertainties, therefore, may originate in the
processes of temporal samplings, error from algorithms, and satellite instruments
themselves. These may affect the accuracy of satellite-derived rainfall products and
may result in a significant error when they are used for various purposes such as
rainfall pattern and variability study. (Mendelsoh et al 2007)
Gridded databases, such as, data generated by atmospheric-ocean coupled global
and regional climate models (e.g., AOGCMs and RCMs), and reanalysis data such as
the National Center for Environmental Prediction – National Center for Atmospheric

13

Research (NCEP-NCAR) are a viable additions and/or alternatives to alleviate the
limitations of limited and inconsistent data, missing information and spatial bias
resulting from the uneven and unrepresentative spatial domain. The reanalyses are
diagnostic atmospheric models, which are used “in concert with observations via data
assimilation”. The reanalysis data are advantageous because they are based on the
AOGCMs with a fixed dynamical core, physical parameterizations and data
assimilation system. (Tarana et al 2011)
A reanalysis is generally a model-run constrained by observations. The space and
time resolution of the data generated through these reanalyses projects are
independent of the number of observations, since the areas void of observations are
filled with dynamically and physically consistent model-generated information.
Although they provide datasets for any period of time, it is evident that their usefulness
crucially depends on the quality and distribution of the observations in time and space.
At the same time, it is important to note that to date, this is the most accurate way of
interpolating data in time and space as well as a superior way to obtain consistency
between different atmospheric variables. It is also more representative because it
provides an opportunity to eliminate local effects, such as those caused by
urbanization (Kalnay and Cai, 2003).
When enough observations are available, the model is more forced to follow the
observed variability rather than its own built-in variability. Assuming that different
datasets have their own variability, there may be instances where at least one of the
reanalyses products do not represent the correct scenario. Comparing results from at
least two reanalyses offers a more correct evaluation of their performances. If the
results agree, the observational constraint is considered large enough to force the
models to follow the real variability of the atmosphere. Alternatively, a difference in the

14

results indicates weak constraints set for that spatio-temporal domain, thereby
indicating that at least one of the products does not represent the correct variability.
So, a difference in two reanalyses products indicates lack of spatial coverage. (Sterl
2004)
Radar data (reanalysis data) is considered as superior to rain gauge data with regards
to describing the spatio-temporal characteristics of rainfall (accumulated rainfall and
rainfall extremes). In addition, considering the high costs of maintenance of land-
based meteorological stations, the density of rain gauge network will decrease in the
future and radar (reanalysis) data will play important role in providing complimentary
information about precipitation. (Kuleshov 2017)
WEATHER STATION OBSERVED DATA

Weather stations measure a large variety of different meteorological parameters,
including air temperature, atmospheric pressure, rainfall, wind speed and direction,
humidity. Manual observations are taken at least once daily, while automated
measurements are taken at least once an hour. In some countries the observations
are taken at intervals which may be every day, or once every month or seasonally.
Normal weather stations are composed of; a thermometer for measuring air and sea
surface temperature, a barometer for measuring atmospheric pressure, hygrometer
for measuring humidity, anemometer for measuring wind speed, pyranometer for
measuring solar radiation and a rain gauge for measuring liquid precipitation over a
set period of time. (Bell et al 2018). An example of a weather station is shown in Figure
1 below

15

Figure 1: example of an automatic weather station
Source: https://www.jma.go.jp/jma/en/Activities/amedas/amedas.html
Figure 1 shows example of weather station with sensors and instruments that
measures the climate variables
Advantages of a station observed data
Some of the advantages of station observed data is that it provides real-time
continuous measuring of climate variables on a daily basis, great accuracy since the
record the conditions right at the ground and generally more reliable. In addition there
can be adjustable sampling intervals for different parameters. Also the meteorological
data can be easily monitored. The archived data can also be accessed locally or
remotely. (Mendelsoh et al 2007)

16

Limitations of a station observed data
Despite the advantages mentioned above, station observed data is limited in areal
representation, this is because the station records only the conditions around the
sensors or instruments, and the area under influence is a radius of about 3-5 km
around the sensor site. Spatially it is difficult to cover all areas of interest, especially
those in mountainous and high latitude regions.
Due to various limitations such as resources, personnel and sometimes difficult
weather conditions the data can have missing gaps and sometimes the time series
data may not be sufficiently long. In general it may not be consistent and of a good
enough resolution to allow use in quality controlled analyses. (Mendelsoh et al 2007)
REANALYSIS DATA
Reanalysis a systematic approach to produce data sets for climate monitoring and
research. Reanalysis datasets are created via an unchanging (“frozen”) data
assimilation scheme and model(s) which ingest all available observations every 6-12
hours over the period being analysed. This unchanging framework provides a
dynamically consistent estimate of the climate state at each time step. The only
component of this framework which varies is the sources of the raw input data. (Hylke
et al 2017).
A reanalysis project involves the reprocessing of observational data spanning an
extended historical period. “It makes use of a consistent modern analysis system, to
produce a dataset, which to a certain extent can be regarded as a “proxy” for
observation with the advantage of providing coverage and time resolution often
unobtainable with normal observational network. It is generated with a data
assimilation system combining observations with a numerical weather prediction

17

model. The reanalysis consequently provides a physical picture of the global climate
over a period during which observational data are available. (Nkiaka et al 2017).
A climate reanalysis gives a numerical description of the recent climate, produced by
combining models with observations. It contains estimates of atmospheric parameters
such as air temperature, pressure and wind at different altitudes, and surface
parameters such as rainfall, soil moisture content, and sea-surface temperature. The
estimates are produced for all locations on earth, and they span a long time period
that can extend back by decades or more. Climate reanalyses generate large datasets
that can take up many terabytes of space. (Kuleshov 2017).
The National Centers for Environmental Prediction (NCEP) is involved with two global
reanalysis projects in joint ventures with other organizations. The first is the
NCEP/NCAR Reanalysis (Reanalysis-1), a global reanalysis of atmospheric data
spanning 1948 to present. The second project is the NCEP/DOE Reanalysis
(Reanalysis-2) project, a global reanalysis of atmospheric data spanning 1979 to
present. (Hylke et al 2017)
Many data sources were used in the generation of both reanalyses. Numerical weather
prediction (NWP) requires input of meteorological data, collected by satellites and
earth observation systems such as automatic and manned stations, aircraft, ships and
weather balloons. Assimilation of this data is used to produce an initial state of a
computer model of the atmosphere, from which an atmospheric model is used to
forecast the weather. These forecasts are typically:
-Medium-range forecasts, predicting the weather up to 15 days ahead
-Monthly forecasts, predicting the weather on a weekly basis 30 days ahead
-Seasonal forecasts up to 12 months ahead

18

Both reanalyses are a global set of gridded weather data at a 2.5 degree by 2.5 degree
horizontal resolution. The main difference between these two global reanalysis
projects is the starting date of their period of records. The year 1979 was chosen as a
beginning date with Reanalysis-2 as it coincides with the date of modern satellite
weather models. Reanalysis-1 begins in the year 1948, and the data input pattern,
better known as data assimilation, changes over the course of this reanalysis, making
it an inconsistent (though still scientifically valid) reanalysis record due to there being
no satellite ingest in the early part of the Reanalysis-1 dataset. (Hylke et al 2017).
The different models under Reanalysis 2 data and the periods they cover are
described below:
ERA-Interim
ERA-Interim is the latest global atmospheric reanalysis produced by the European
Centre for Medium-Wave Forecasts (ECMWF) and covers the period from 1 January
1979 to present day. The core component of the ERA-Interim data assimilation system
is the 12-h 4D-variational analysis scheme of the upper-air atmospheric state, which
is on a spectral grid with triangular truncation of 255 waves (corresponding to
approximately 80 km) spatial resolution and a hybrid vertical coordinate system with
60 vertical levels. (Nkiaka et al 2017).
ERA-20CM (Jan 1900 – Dec 2010) Final
ERA-20C is ECMWF’s first atmospheric reanalysis of the 20th century, from 1900-
2010. It assimilates observations of surface pressure and surface marine winds only.
It involves a coupled Atmosphere/Land-surface/Ocean-waves model which is used to
reanalyse the weather, by assimilating surface observations. The ERA-20C products
describe the spatio-temporal evolution of the atmosphere (on 91 vertical levels,

19

between the surface and 0.01 hPa), the land-surface (on 4 soil layers), and the ocean
waves (on 25 frequencies and 12 directions).
ERA-Interim/Land
ERA-Interim/Land is a global reanalysis of land-surface parameters from 1979-2010
at 80 km spatial resolution. It preserves closure of the water balance and includes a
number of parameterisations improvements in the land surface scheme with respect
to the original ERA-Interim dataset, which makes it suitable for climate studies
involving land water resources.
CERA-SAT (Jan 2008 – Dec 2016) (New)
CERA-SAT is a reanalysis dataset spanning 8 years between 2008 and 2016. It has
been produced within the scope of the ERA-CLIM2 project as a proof-of-concept for a
coupled reanalysis with the full observing system available in the modern satellite age.
It accounts for model- and observational errors and can be used to infer information
on the uncertainty of the analysed fields. (Nkiaka et al 2017).
Advantages of Reanalysis data
Reanalyses incorporate millions of observations into a stable data assimilation system
that would be nearly impossible for an individual to collect and analyze separately,
enabling a number of climate processes to be studied. (Nkiaka et al 2017). It is
essential for forecast products predicting anomalous or extreme weather. Forecasts
of severe weather events allow appropriate mitigating action to be taken and
contingency plans to be put into place by the authorities and the public. The increased
time gained by issuing accurate warnings can save lives, for instance by evacuating
people from a storm surge area. Authorities and businesses can plan to maintain
services around threats such as high winds, floods or snow. It enables estimates of

20

quantities and physical processes that are very difficult to observe directly, such as
vertical motions, surface heat fluxes and waves. (Kalnay and Cai, 2003).
Inspite of the many advantages of using reanalysis data sets, there can be
inhomogeneities in the observing system and biases in the numerical models which
may lead to spurious changes and trends. (Nkiaka et al 2017). However these
limitations can be corrected by intercomparing the output from different reanalyses,
and comparing reanalysis output to independent observations
THE WATER CYCLE INTEGRATOR (WCI)

WCI portal is an open source project built by Plymouth Marine Laboratory’s Remote
Sensing Group. The project started in 2014. This portal integrate available earth
observations, in-situ datasets and models, in order to construct a consistent global
water resources reanalysis.
The WCI portal takes data that you select and plots it on a map to help you analyse.
This information is vital to provide support for water related actions and planning in an
informed decision making process. The tool was developed as a way of addressing
some of the data challenges encountered in trying to address water resource problems
and for the following reasons (http://www.earth2observe.eu/)
-The availability and quality of water resources are still not fully accounted for in
different regions of the world.
-Many countries lack basic information such as observations of the required hydro-
meteorological variables.
-A consistent global reanalysis dataset of water resources, robust and long enough to
capture climate variability is currently lacking

21

-Ideally station observed data is ideal for use in measuring climate variables such as
precipitation, however due to the limitations above it is better to use the reanalysis
data because it provides more spatial coverage and has high resolution

22

CHAPTER TWO
MATERIALS AND METHODS

THE STUDY AREA

Mahalapye is located in the Central District in Botswana located at the Geographical
coordinates of latitude: -23.067 and longitude: 26.833. The climate of Mahalapye is a
local steppe climate. During the year, there is little rainfall in Mahalapye. According to
Köppen and Geiger, this climate is classified as BSh (arid and desert climate). In
Mahalapye, the average annual temperature is 20.4 °C. Precipitation average is 452
mm. (Lohmann et al 1993). Figure 2 below shows the study area within the longitudes
and latitudes. The people in this area are involved in farming, they grow crops and
rear animals.

23

Figure 2: shows the study area of Mahalapye
Source: ArcGIS
DATA AND METHODS

The rainfall data was obtained from the Mahalapye meteorology station. The R
software was downloaded from this link, https://cran.r-project.org/bin/windows/base/
and it was used to calculate the drought indices. R is a programming language and
environment used for statistical computing and graphics. The Mahalapye data was
sorted into the variables required by the calculator before putting it into the calculator.
Thereafter it was put into the calculator and the calculator produced the output in the
form of spreadsheet data and graphs for both the SPI and SPEI

24

The reanalysis data were obtained from the Earth Observation portal. The portal
integrate available earth observations, in-situ datasets and models, in order to
construct a consistent global water resources reanalysis. It takes data that you select
and plots it on a map to help you analyse
Calculation procedure and the interpretation of SPI by the R software
The method used for SPI computation was developed by McKee et al. (1993) and
Edwards and McKee (1997) to study relative departures of precipitation from
normality. It uses monthly precipitation values at different time scales (1, 3, 6, 12, 18,
and 24 months. However in this study we used a timescale of 12 months. It is often
called the z score, is the number of standard deviations from the mean at which an
event occurs. Thus, 12 month SPI value provides a comparison of accumulated
precipitation over that specific 12 month period with the mean precipitation total for the
same annual period as calculated over the full study period. For precipitation, high
positive values correspond to wet sequences and high negative values correspond to
drought periods.

25

CHAPTER THREE
RESULTS

year Annual Average Temperature(Degrees Celsius)
1980/81 28,1
1981/82 29,6
1982/83 30,4
1983/84 29,4
1984/85 29,0
1985/86 29,3
1986/87 30,0
1987/88 28,4
1988/89 28,2
1989/90 29,3
1990/91 29,0
1991/92 30,8
1992/93 29,6
1993/94 28,7
1994/95 28,6
1995/96 28,3
1996/97 29,1
1997/98 30,9
1998/99 29,4
1999/00 27,3
2000/01 29,1
2001/02 29,8
2002/03 30,3
2003/04 28,4
2004/05 30,1
2005/06 28,4
2006/07 30,7
2007/08 28,1
2008/09 29,0
2009/10 29,1
2010/11 28,7
2011/12 29,8
2012/13 29,6
2013/14 28,7

Table 3.1: shows the annual average temperature recorded at Mahalapye weather
station

26

Year jul aug sept oct nov dec jan feb march april may jun AVERAGE
1979/80 0 48,5 0 34,9 54,6 26,1 120,2 138,4 64,3 10,7 0 0 41,5
1980/81 0 0 6,3 20,2 104,8 63,1 135,7 137,3 98 36,2 0 0 50,1
1981/82 0 11 0 19,2 145,7 76,4 50,9 12,2 65,1 44,8 0 0 35,4
1982/83 1,4 0 2,4 96 48,9 85,9 26 15,4 13,1 14,9 1,9 3 25,7
1983/84 1,4 2,1 0,2 17,4 81,8 58,5 71 24,2 257,2 1,4 1,4 0,2 43,1
1984/85 11,5 0 0,3 21,3 47,6 29,5 79,8 32,6 36,3 0,6 0 0 21,6
1985/86 0,1 1,9 0 46,6 15 31,3 3,1 44,3 22,6 86,1 0 5,5 21,4
1986/87 0 0 12,5 60,8 148,7 41,1 32,4 11,4 17,6 1,6 0 0 27,2
1987/88 0 0 21,6 1 57,3 83,3 57,1 447 119,7 46,7 0 0 69,5
1988/89 0 0 12,1 25,1 32,7 75,9 51,3 149,1 37,9 40,9 0 28,5 37,8
1989/90 0 0,6 0 24,3 63,4 74,2 62,9 102,6 129,5 21,8 7,5 0 40,6
1990/91 0 0 2,1 32,3 25,8 83,5 173,4 112,1 107,2 0 2,5 22,6 46,8
1991/92 0 0 4,7 23,8 26 26,5 90,4 40,1 27,5 2,7 0 0 20,1
1992/93 0 0 6,4 85 60,1 124,7 96,8 56 38,4 8,3 0,2 0 39,7
1993/94 7,2 0 19,8 32,4 67,5 72 50,9 51,8 28,2 2,6 0 0 27,7
1994/95 0 0 0 11,1 136,5 30,5 49 161,3 67,4 30,3 16,8 0 41,9
1995/96 0 0 1,3 7,9 52,6 327,4 84 102,7 0,6 2 9,2 0 49,0
1996/97 0 0 1,2 14,6 126,8 44,9 245,8 7,2 70,7 5,6 31,6 0 45,7
1997/98 0 0 8,1 17,8 108,4 108,4 87 4,2 66,9 0 0 0 33,4
1998/99 0 0 0 35,7 92,2 147,3 18,9 5,7 14,8 0,3 7,8 0 26,9
1999/00 0 0 0 1,3 51 146,6 122,5 262,6 100,6 24,6 8,4 8,7 60,5
2000/01 0 0 0 11 24,1 35,8 8,5 127,3 36,5 15,9 6,6 5,2 22,6
2001/02 0 0 0 8,7 199,3 50,7 5,7 19,8 7,4 40,3 24 0,5 29,7
2002/03 0 12,5 8,9 19,2 2 109,5 26,4 85,7 22,7 0 0 31,9 26,6
2003/04 0 0 4,9 22,4 67,4 32 127,8 75,9 128,6 34 0 0 41,1
2004/05 0 0 0 40,2 56,8 127,9 79,4 34,2 30,6 94,8 0 0 38,7
2005/06 0 0 0 0 77,7 94,7 113,1 129,6 119,6 0,1 29,3 0 47,0
2006/07 0 0 0 0,7 23,2 105,4 77,2 10,3 56,8 18,1 0 0 24,3
2007/08 0 0 60,5 54,5 127,5 221,9 129,1 50,5 24,9 3,5 4,8 0 56,4
2008/09 5,2 0 0 14 105,5 45,8 230,2 57 91 0 30,1 92,7 56,0
2009/10 1 0 14,8 83,5 57,2 77,2 42,9 15,4 16,8 191,8 3,3 0 42,0
2010/11 0 0 0 100,1 38,7 28,7 81,5 76 3,8 35,8 0 1,6 30,5
2011/12 1,4 0 0 10,7 76,4 67,8 31 9,3 16,7 1,5 0 0 17,9
2012/13 0 0 0 82,7 61,4 95,6 69 0,7 6,5 28,5 0 0 28,7
2013/14 0 0 0 14,5 57,7 149,2 107,8 25,3 156,5 23,7 0 0 44,6

Table 3.2: shows the monthly precipitation and annual average precipitation recorded
at Mahalapye meteorological station

27

Table 3.3: SPI indices from station data

JanFebMarAprMayJunJulAugSepOctNovDec
1980NANANANANANANANANANANA0.462635
19810.6126764660.6809214810.7965263280.9617303180.9472146450.9374720.9320660.9878160.9771210.9252161.1412471.162717
19820.807745925-0.017468035-0.192540657-0.142867343-0.135827553-0.13103-0.11886-0.19219-0.186160.292849-0.35076-0.28834
1983-0.532843615-0.554801134-0.810697962-1.098748955-1.056557529-1.01384-1.00446-0.98176-1.03482-1.76504-1.45242-1.68668
1984-1.442562256-1.4970741730.5526501440.4822134980.4740263830.4538660.5118820.4983790.5084090.496130.2877070.089621
19850.1526878270.242515046-1.440751383-1.499839517-1.479894698-1.45897-1.55639-1.53045-1.58367-1.32969-1.65427-1.60313
1986-2.78173076-2.897717063-2.514186595-1.600871686-1.565386962-1.48785-1.47584-1.48697-1.4144-1.27345-0.23266-0.17335
19870.008437547-0.251087205-0.196706157-0.861041025-0.839988477-0.87048-0.86219-0.85767-0.81308-1.32603-2.29887-1.78262
1988-1.7511072931.6238982081.8668700322.1298770182.0925653212.0674472.0534792.0458732.0581142.0904591.9885241.887679
19892.0970454910.465933827-0.095790763-0.144986572-0.137905430.0574930.0587530.063324-0.02126-0.049020.1455820.118216
19900.203991023-0.1334057460.5130502540.4056367520.4478961310.2702880.2699360.2660470.2819690.3014240.0608330.104563
19910.8310254990.9953446190.7379695590.6270594340.5895836430.7147760.7110570.7090360.7396620.6508260.6505170.291949
1992-0.273800964-0.943803495-1.41982033-1.444023674-1.43549985-1.63387-1.61979-1.61212-1.64735-1.06863-0.79765-0.09319
1993-0.0689601110.0589874340.1631391080.1998248260.2014623260.2017320.2469560.2468650.333654-0.031190.012608-0.35018
1994-0.781099755-0.898819259-0.775369192-0.852688057-0.833435124-0.81927-0.86951-0.86495-1.06431-1.24744-0.68469-1.00363
1995-1.163481533-0.2670344980.0876158650.2674880170.3717272810.3697110.3686060.3680090.3816740.330093-0.217431.426151
19961.7870173251.647544711.0555363080.926239520.8701162610.8614090.8565790.8539520.8732750.8657931.255331-0.43835
19970.6399909350.00414440.4862665560.5185213930.645464810.6397740.6366220.6349110.6886910.6672690.5600640.877416
1998-0.058035163-0.080395571-0.04720462-0.093002459-0.30887708-0.30176-0.29777-0.2956-0.36852-0.25645-0.37722-0.11669
1999-0.675977763-0.728429683-0.969552526-1.00316461-0.913636968-0.8984-0.8899-0.88526-0.91651-1.21009-1.60741-1.58184
2000-0.8041387471.1209857091.4103675951.573060951.549562521.5737221.5634941.5579271.5985181.5830151.4568760.849487
20010.221204655-0.894058143-1.22769189-1.353766871-1.339883862-1.35134-1.33941-1.3329-1.37731-1.38868-0.036810.046373
20020.022053136-0.899772042-0.930263905-0.761104728-0.60514537-0.63007-0.6236-0.52607-0.47922-0.41427-2.21451-1.5492
2003-1.545477083-1.015489831-0.669676625-1.033604706-1.22035516-0.93067-0.92193-1.02123-1.0914-1.06097-0.54933-1.15582
2004-0.436453908-0.5628074720.2943708650.5117725780.5060377530.308830.3081860.3078410.2807010.3597930.292020.811817
20050.6082164760.380761772-0.2879691240.1218506090.1237248250.1250380.1257870.12620.124681-0.16642-0.03151-0.26251
2006-0.0558531350.6674499691.1210224220.6168533530.7787416690.7712610.7671140.764860.8003720.7619940.4589570.49579
20070.313827389-0.62011036-1.01434471-0.904757244-1.133591225-1.1154-1.10526-1.09972-0.65923-0.273150.3846970.997827
20081.4119265231.8177883741.3819884361.3449223931.3474535741.3323381.3498291.3451511.0621670.7951840.67102-0.4575
20090.221891970.3041276270.6884400990.6843036970.8201236181.3035931.2742931.2699281.3779551.6563361.4285481.522804
20100.6374560190.415725936-0.0903655451.082702810.9196166340.3758160.3686060.3680090.2807010.3525380.23563-0.08799
20110.7029721261.6717082291.3729950680.7520256680.7226397680.7340550.7458890.7437230.7603640.2565350.9132861.407267
20121.0143482760.187169060.299965084-0.157723092-0.15039342-0.16739-0.18361-0.18191-0.192490.6519620.4761570.251254

28

Table 3.4: SPEI indices for station data

JanFebMarAprMayJunJulAugSepOctNovDec
1980NANANANANANANANANANANA1.025376
19811.1747682511.3963958311.417167621.4786030111.4665981141.4475931.4224121.4754191.4688071.5263311.5874611.604427
19821.3528400490.6475588450.2473912740.3586244560.3108448040.2981140.3047930.2012330.1573530.509592-0.02397-0.14325
1983-0.490838958-0.63750983-0.956145454-1.280729023-1.246395307-1.25059-1.25451-1.19351-1.29367-1.84552-1.62362-1.68187
1984-1.543354231-1.6437190460.371411910.4227227880.4146437330.4402630.5065340.4495910.4499090.3531450.2974320.041024
19850.2105352560.457012281-1.109771343-1.10596169-1.047075343-1.05662-1.11531-1.12154-1.05497-0.86083-1.14225-1.1046
1986-1.759489814-1.989584657-1.767185389-1.258618799-1.228984563-1.15993-1.15863-1.19775-1.13517-0.963040.0035-0.02119
19870.201546095-0.181451835-0.200883271-0.870500117-0.898560023-0.90294-0.87403-0.78609-0.72375-1.11777-1.75987-1.56067
1988-1.6586306971.4764750331.7863826051.9356083871.9631715991.9301871.9122211.8848251.8882221.9175251.9330591.988246
19892.1527387991.1305747130.5395609810.5920976020.5329424530.6600660.6650720.6201850.539620.4929160.6105480.468292
19900.6341463370.4369442940.9232839640.7486692190.7926788440.6365950.6018830.6904080.7382550.669360.2540020.183982
19910.8669681680.9767700990.7928426850.7562284650.6716331220.7969330.8347380.8181210.7801520.6752930.6876450.413407
1992-0.236330743-1.093620181-1.594897077-1.800027912-1.727087018-1.85833-1.86978-1.83105-1.94354-1.59022-1.24356-0.81839
1993-0.760866202-0.462840084-0.1656536170.029222211-0.037264657-0.011580.013454-0.01560.123118-0.18278-0.11001-0.42605
1994-0.630652992-0.649589785-0.640041853-0.647541408-0.545200924-0.49179-0.44284-0.4041-0.47018-0.45805-0.11146-0.48072
1995-0.752562959-0.1728487420.1274307450.3723572150.4606967010.4523290.4046230.3500570.3009240.055917-0.526971.244569
19961.6241794131.6849243061.2917003451.1926228461.1262277171.1030091.1307051.1686421.2270621.2247851.532035-0.00108
19971.0113614850.457222790.8213460130.871393860.9901732780.9810970.9538420.9191030.977471.0681330.8982511.083373
19980.2532337080.245868402-0.032054031-0.256838432-0.565820396-0.58877-0.61762-0.566-0.66681-0.60299-0.63692-0.26002
1999-0.74662143-0.790248068-0.952000218-0.865625393-0.776294676-0.73701-0.71745-0.72664-0.63342-0.81311-1.13245-1.23677
2000-0.5600658691.3184184461.6151532911.7255167561.7480423921.7569971.7712451.7745351.7773141.7117011.6840281.232002
20010.598764731-0.329989055-0.734771437-0.861518675-0.899404425-0.90617-0.89225-0.95699-0.97753-1.000940.1121120.138727
20020.185106469-0.729756142-0.830381029-0.688894706-0.562424686-0.55937-0.57875-0.48615-0.44157-0.4042-1.66109-1.40606
2003-1.555019579-1.375762192-1.039035892-1.310305127-1.404455044-1.18736-1.17657-1.18471-1.25021-1.2619-0.88458-1.46936
2004-0.917269245-1.074439255-0.1129903680.181189960.146538877-0.05001-0.04821-0.14763-0.13476-0.0437-0.237260.332886
20050.027340088-0.360699476-0.999086824-0.487631169-0.511137455-0.57355-0.6107-0.63713-0.80883-1.12802-0.8713-1.02513
2006-0.873248425-0.1000195940.6509617180.0927095630.3699164140.4403670.4314710.5550240.6951410.6757370.4239240.327356
20070.166725432-0.915386545-1.397847805-1.358981654-1.527220887-1.52762-1.47727-1.52319-1.31458-0.78555-0.125560.672797
20081.1014617611.5486771781.3240486231.3036889461.2558580641.2304341.2294621.1884260.9838510.6127180.416525-0.86087
2009-0.389183556-0.3408591660.1722303750.1253196240.2893811290.8097820.8275660.8623210.9109471.2310351.0641951.179112
20100.2882029870.027345933-0.5559419520.7426612680.5316815420.025833-0.004540.054151-0.02976-0.0013-0.25471-0.52438
20110.3692377751.3946011561.1658717150.5296172750.5051333550.5139310.555260.5746510.5904430.1951770.8159061.309047
20120.837813466-0.1295179810.062351484-0.37240568-0.382493689-0.40362-0.46359-0.52369-0.467080.417050.204841-0.07554

29

Table 3.5: SPI indices for reanalysis data

minmaxmeanstdmedian
1983-0.251780.045413-0.085790.077133-0.06822
1984-1.08740.510015-0.379490.398313-0.4872
1985-1.78530.818313-0.676770.819132-0.85814
1986-2.244920.517493-1.108710.577096-1.04872
1987-1.665261.4495670.1747030.7394160.17143
19880.9517753.2260952.1472210.6197962.136434
1989-0.674851.3384710.1128030.4728260.053405
1990-1.931750.705102-0.725650.634515-0.75307
1991-0.465181.0328580.1415690.3697510.151074
1992-3.54391-1.28797-2.251750.519884-2.20805
1993-2.190730.957659-0.746590.743411-0.86557
1994-1.282350.670857-0.544760.491025-0.65296
1995-1.28751.810616-0.009580.670933-0.1316
1996-0.356911.6356921.0586720.4239411.143254
1997-0.41392.2959780.7987140.791080.624609
1998-1.609450.828139-0.056850.6615840.130201
1999-0.376990.8497370.282980.2564250.267822
20000.6521532.5441661.3375240.4202621.317758
20010.3273621.5693211.0168890.2652561.029796
2002-1.070240.9651960.0692990.4764-0.00869
2003-1.80982-0.19933-0.953780.431877-0.92913
20040.0880830.4719980.3146110.0934620.334142
2005-0.514980.182789-0.166540.186967-0.14867
20060.2324131.5396370.949240.3456260.996821
2007-1.888230.248398-0.509520.628066-0.25164
2008-0.423240.191855-0.069790.13051-0.06306
2009-0.327850.5682270.132580.2326640.101265
20100.1596671.3670360.9622970.2908330.952706
20110.0364041.0108950.66490.2301620.645092
2012-0.99224-0.10599-0.511210.201531-0.49408

30

Table 3.6: SPEI indices for reanalysis data

minmaxmeanstdmedian
1983-1.72043-0.41982-0.944780.267877-0.95103
1984-1.74785-0.01869-0.848960.32623-0.79937
1985-1.602110.686395-0.5860.775982-0.72068
1986-1.625310.437256-0.870940.434834-0.85443
1987-1.430240.406798-0.608920.421164-0.69307
19880.8497952.4204371.9591120.4025052.115896
1989-0.497011.1907440.1640340.4151870.126756
1990-1.01690.62466-0.36920.386352-0.42585
1991-0.525020.540685-0.127420.253483-0.14921
1992-2.77899-1.47033-2.252870.362231-2.35004
1993-1.400680.109891-0.765570.367319-0.8275
1994-0.835190.601448-0.237060.365768-0.27382
1995-1.65205-0.39189-1.278960.301967-1.34176
19960.2844971.8924181.2993770.3529711.408617
1997-0.063222.2567930.9098720.5499260.853174
1998-1.117270.51745-0.168260.458551-0.18666
19990.0369710.8535340.4724970.1995250.438955
20001.5619912.3907322.0043140.1942972.022897
20010.6668771.551581.0658430.205671.066449
2002-0.919860.9419580.0111640.4194890.000958
2003-1.72719-0.80235-1.323180.25818-1.35909
2004-0.137290.4126870.1395450.1125780.137446
2005-0.90219-0.18782-0.4940.180554-0.45764
2006-0.123751.586511.0443760.3651811.104223
2007-1.654210.09733-0.635030.513853-0.52264
20080.1135640.6494140.3843670.1294180.386347
2009-0.147750.7096120.4240160.2432540.510382
20100.055261.2844050.8430490.2972170.885869
20110.1850081.0751990.7323890.2045960.740204
2012-1.17793-0.25141-0.714290.237622-0.71583

31

Graph 1: shows the SPI output from Mahalapye station data

Graph 2: shows the SPI output from reanalysis data

32

Graph 3: SPEI output from Mahalapye station data

Graph 4: SPEI output from reanalysis data

33

Range Description
2.0+ Extremely wet
1.5 to 1.99 Very wet
1.0 to1.49 Moderately wet
-0.99 to 0.99 Near normal
-1 to -1.49 Moderately dry
-1.5 to -1.99 Severely dry
-2 and less Extremely dry

Table 3.7: the SPI index values

YEAR STATION DATA REANALYSIS DATA
1986 -3 -2
1988 +2 +2
1992 -1.5 -2.4
2001-2005 Average = -1 Average= +1
2008-2013 Average= +1.5 Average= +1

TABLE 3.8. INTERPRETATION OF VALUES OF SPI

34

YEAR STATION DATA REANALYSIS DATA
1982 +1.5 -1.5
1988-1990 Average= +1 Average= +2
1992 -1.8 -2.3
1999 -1 +2
2008-2013 Average= +1 Average= +0.8

TABLE 3.9. INTERPRETATION OF VALUES OF SPEI
The reason the above years were chosen is because the peaks of the graphs for the
two data models were different denoted by the highest peaks from other peaks the
chosen years

35

CHAPTER FOUR
DISCUSSION

As for the standardized precipitation index, from table 1, the results shows that the
year 1986 had more drought which is described by the values from (McKee et al 1993)
which shows the year was extremely dry. The results shows that the year 1992 was
hit by drought as it is represented by the high negative values of the index from both
the station observed data and reanalysis data of -1.5 and -2.4 respectively. For some
years the values were around zero which depicted nearly normal condition. According
to McKee, the drought is very extreme when the SPI values are below -2 and this
shows that on average the Mahalapye experienced moderately dry weather.
The results shows a bit of a difference between the station data output and reanalysis
data and the difference is brought by the fact that the satellite data represent areal
rainfall, while the gauge data represent point rainfall and that this point needs to be
taken into consideration in making the comparisons between the two sources of
precipitation data. The other reason is because the satellite data is not sensitive to
topographic influences on precipitation and gauge data is commonly not adequate to
completely quantify spatial variations due to topographic influences, hence there is a
difference between the two data outputs. (Hughes et al 2006).
In both data sources the results shows that from 2008 to 2013 the area experienced
moderately wet conditions which shows the rainfall has been high/good and low
drought rate
As for the SPEI, standardised precipitation evapotranspiration index, the results shows
that only the year 1992 was hit by extreme drought which yielded the values of -1.8
and -2.3 from the station data and reanalysis data respectively. The results shows a

36

small difference between the station data and the satellite data. However the
difference between the two data sources is because the station data represent the
point rainfall and temperature whilst satellite data represent areal rainfall and
temperature.
The measure difference in the values of both the SPI and SPEI in both data sources
type is because methods used by the satellites to make rain estimates are indirect.
Ground-based satellites radar systems provide useful information about rainfall over
a large area, so this means that the methods used by radar to make quantitative
estimates of rainfall suffer from the fact that they observe rain from a distance.
Whereas rain gauges, on the other hand, measure rainfall directly, but only over a
small area. The errors rain gauges make are fairly well understood, and so, except for
their limited coverage, they can be ideal for making rainfall estimates. (Bell and Kundu
2018)
Both the SPI and SPEI in both data sources shows that year 1992 was affected by too
much drought depicted by high negative values of -1.5 and -2.4 in SPI and -1.8 and –
2.3 in SPEI
However the difference between the standardised precipitation index and the
standardised precipitation evapotranspiration index is because the SPI uses only the
rainfall data/ amount to calculate the drought indices whereas the SPEI uses both the
rainfall data and temperature data to calculate the indices. So as such if the
temperature values are high it means the indices values will be lower which depicts
dry and extremely dry conditions and this means there is very high drought

37

CHAPTER FIVE
CONCLUSION

The data outputs for all the drought indicators (SPI and SPEI) shows that the year
1992 was hit by an extreme drought represented by the values ranging from -1.5 to –
2.4 of the SPI index description. The results shows that there is a small difference
between the reanalysis data and a weather station observed data and the difference
is brought up by the difference in areal cover of the two data sources/models. However
according to the result, it shows that the reanalysis data can be coupled with a station
observed data to produce a better and accurate results.

38

REFERENCES

H. B. Lakew, S. A. Moges, D. H. Asfaw, (2017). Hydrological Evaluation of Satellite
and Reanalysis Precipitation Products in the Upper Blue Nile Basin: A Case Study of
Gilgel Abbay
R Mendelsoh, P Kurukulasuriya, A Basist, F Kogan, C Williams. (2007). Climate
analysis with satellite versus weather station data: climatic change, 81(1), 71–83. DOI:
10.1007/s10584-006-9139-x
Y Kuleshov, (2017). Use of Remote Sensing Data for Climate Monitoring in WMO
Regions II and V (Asia and the South-West Pacific)
Hylke E. Beck, I. J. Albert, M. van Dijk, V Levizzani, J Schellekens, G. Diego, B
Martens, and A de Roo. (2017). MSWEP: 3-hourly 0.25 global gridded precipitation
(1979–2015) by merging gauge, satellite, and reanalysis data. DOI: 10.5194/hess-21-
589-2017
Tarana A. Solaiman, Slobodan P. Simonovic. (2011). Assessment of Global and
Regional Reanalyses Data for Hydro-Climatic Impact Studies in the Upper Thames
River Basin
L. Wang, C. Thomas, Peterson, J. Lawrimore, M. Brunet-India, R. Cerveny, C. Donlon,
F. Driouech, A. Wan Hassan, R. Hollmann, D. Schwartz9, Z. Zhang. (2008).
Monitoring the Earth’s Climate. 57 (2)
G Tesfaye Ayehu, T Tadesse, B Gessesse, T Dinku. (2018). Validation of new satellite
rainfall products over the Upper Blue Nile Basin, Ethiopia, DOI.org/10.5194/amt-11-
1921-2018

39

E Nkiaka, N. R. Nawaz, J C. Lovett. (2017). Evaluating Global Reanalysis Datasets as
Input for Hydrological Modelling in the Sudano-Sahel Region
E. Kalnay, and M. Cai, (2003). Impact of urbanization and land use change on climate.
Nature, 423: 528-53
Beljaars, A., Boussetta, S., Brun, E., Cloke, H., Dee, D., Dutra, E., Pappenberger, F.,
de Rosnay, P., Stockdale, T., and Vitart, F. (2015): ERA-Interim/Land: a global land
surface reanalysis data set, Hydrol. Earth Syst. Sci., 19, 389-407, doi:10.5194/hess-
19-389-2015. http://www.hydrol-earth-syst-sci.net/19/389/2015/hess-19-389-
2015.html
T.B McKee, N.J. Doesken, J. Kleist, (1993): The relationship of drought frequency and
duration to timescale. In: Proceedings of the Eighth Conference on Applied
Climatology, Anaheim, California, 17–22 January 1993. Boston, American
Meteorological Society, 179–184.
G.W. Kite, P. Droogers, (2000) Comparing evapotranspiration estimates from
satellites, hydrological models and field data, J. Hydrol. 229 (2000) (1–2), pp. 3–18
D. A. Hughes, L. Andersson, J. Wilk, H. H. G. Savenije, (2006). Regional calibration
of the Pitman model for the Okavango River. J. Hydrol., in preparation
L. Bell, K. Kundu, (2018). Comparing Satellite Rainfall Estimates with Rain-Gauge
Data: Optimal Strategies Suggested by a Spectral Model
U. Lohmann, R. Sausen, L. Bengtssonl, U. Cubasch, J. Perlwitz, E. Roeckner. (1993).
The Koppen climate classification as a diagnostic tool for general circulation models.
(3) 177-193
http://thegpscoordinates.net/gpsimg/mahalapye.png

40

x

Hi!
I'm Mila

Would you like to get a custom essay? How about receiving a customized one?

Check it out