A comparative study of surface air temperature between satellite and model in vietnam from 2004 to 2013

THAI NGUYEN UNIVERSITY UNIVERSITY OF AGRICULTURE AND FORESTRY TRAN THI THU THAO A COMPARATIVE STUDY OF SURFACE AIR TEMPERATURE BETWEEN SATELLITE AND MODEL IN VIETNAM FROM 2004 TO 201

THAI NGUYEN UNIVERSITY

UNIVERSITY OF AGRICULTURE AND FORESTRY

TRAN THI THU THAO

A COMPARATIVE STUDY OF SURFACE AIR TEMPERATURE BETWEEN

SATELLITE AND MODEL IN VIETNAM FROM 2004 TO 2013

BACHELOR THESIS

Study Mode: Full-time Major: Environmental Science and Management Faculty: Advanced Education Program Office Batch: 2014-2018

Thai Nguyen, 10/08/2018

Thai Nguyen University of Agriculture and Forestry

Degree Program Bachelor of Environmental Science and Management

Student name Tran Thi Thu Thao

conducted not only to compare which datasets captures the surface air temperature

in better way, but also to see that during the years of the study climate change has occurred The research performs that AIRS data able to present the effect of topography and climate on the surface air temperature better than ECMWF model Keywords Surface air temperature, atmospheric infrared sounder(AIRS),

ECMWF reanalysis

Number of pages 53

Date of submission 25th September, 2018

ACKNOWLEAGEMENT

In the process of implementing the research topics: " A comparative study of surface air temperature between satellite and model in Vietnam from 2004 to 2013" I got the guide and help, encouragement of the lecturers and the other people I would like to express sincere gratitude to:

First and foremost, I wish to express sincere thanks to the Director Board of

Thai Nguyen University of Agriculture and Forestry, board of managers of International Training and Development Center and all the teachers have

dedicated teaching, the scientific knowledge and the valuable experience for me during the training period, learning in school

I sincerely thank the cooperation between Thai Nguyen University of Agriculture and Forestry and National Central University for giving me a

wonderful opportunity to internship in Taiwan It gives me new knowledge with a thesis to graduate

In particular, I take this opportunity to express profound gratitude and deep

regards to the supervisor Prof Dr Chian Yi-Liu who guided wholeheartedly, to

advise, imparted knowledge and valuable experience to help me during the internship

in Taiwan

Furthermore, I express my appreciation to Assoc.Prof Dr Tran Quoc Hung

for his enthusiastic guidance His detailed instructions and encouraging after I returned from my internship in Taiwan helped me to complete the report

I am also thankful to Mr Chi-Hao Chiu and Ms Tran Huyen Trang for

teaching me various techniques, methods used in satellite and model data and helps me

to analysis the data Without the comments and the motivation from them, my research work was not able to complete well

Finally, I would like to express our gratitude to my family, relatives and friends who have helped me during the internship I would not have been able to complete this thesis without their continuous help and encouragement

Sincerely,

Tran Thi Thu Thao

CONTENTS

LIST OF FIGURES vii

LIST OF TABLE viii

LIST OF ABBREVIATIONS ix

PART I INTRODUCTION 1

1.1.Research rationale 1

1.2.Research’s objectives 3

1.3.Research question 3

1.4.Limitations 3

PART II LITERATURE REVIEW 4

2.1 Surface air temperature 4

2.2 AIRS satellite, ECMWF model and Matlab 9

PART III METHODOLOGY 14

3.1 Data collection 14

3.2 Region selected 15

3.3 Data Analysis 17

PART IV RESULTS 19

4.1 The different between AIRS satellite and ECMWF model about SAT 19

4.1.1 The different between AIRS satellite and ECMWF model about SAT in sub regions 19

Fig 6: AIRS measured and ECMWF model estimated over Tay Nguyen region 23

4.1.2 The difference between three sub regions about monthly SAT by ECMFW model estimated and AIRS measured 27

4.1.3 The different between AIRS satellite and ECMWF model about SAT in

Vietnam 28

4.2 The different between AIRS satellite and ECMWF model about SAT in each season 33

PART V DISCUSSION AND CONCLUSION 37

5.1 Discussion 37

5.2 Conclusion 38

REFERENCES 40

LIST OF FIGURES

Fig 1 Frame word of study 14

Fig 2 Vietnam region with latitude interval 70N-250N and longitude interval 1000E-1100E (from Google earth) 15

Fig 3 Bac Bo region with latitude interval 210N-22030'N and longitude interval 103030'E-116030'E (from Google earth) 16

Fig 4 Tay Nguyen region with latitude interval 120N-14045'Nand longitude interval 1070E-1080E (from Google earth) 16

Fig 5 Nam Bo region with latitude interval 9045'N-10045'Nand longitude interval 105015'E-106015'E (from Google earth) 17

Fig 6 AIRS measured and ECMWF model estimated over Bac bo region 20

Fig 7 AIRS measured and ECMWF model estimated over Tay nguyen region 23

Fig 8 AIRS measured and ECMWF model estimated over Nam bo region 26

Fig 9 ECMWF model estimated over 3 sub regions 27

Fig 10 AIRS measured over 3 sub regions 28

Fig 11 AIRS measured and ECMWF model estimated over study region 31

Fig 12 The different between AIRS satellite and ECMWF model annual SAT by month in each season in Bac Bo region (◦C).a) Hot season; b) Cold season 34

Fig 13 The different between AIRS satellite and ECMWF model about annual SAT by month in each season in Tay Nguyen region(◦C) a) Rainy season; b) Dry season 35 Fig 14 The different between AIRS satellite and ECMWF model about annual SAT by month in each season in Nam Bo region (◦C) a) Rainy season; b) Dry season 36

LIST OF TABLE

Table 1: CC, difference and rsme between AIRS measured and ECMWF model estimated SAT 31Table 2: The average annual SAT of study region and sub regions 32Table 3: CC between AIRS measured and ERA model estimated SAT

in each seasons 33

ABBREVIATIONS

AIRS Atmospheric Infrared Sounder

CC Correlation coefficient

DHS data handling system

ECMWF European Center for Medium-range Weather Forecasts

EOS Earth Observing System

IMHEN Vietnam on Disaster Risk Management and Extreme

Conditions to Promote Climate Change Adaptation IPCC Intergovernmental Panel on Climate Change

JCSDA Joint Center for Satellite Data Assimilation

Matlab MATrix LABoratory

NWP Numerical weather prediction

RCPs Representative Concentration Pathways

RMES Root mean square error

SAT Surface air temperature

t2m 2 meter temperature

PART I INTRODUCTION 1.1 Research rationale

The temperature of the air near the surface of the Earth, usually determined by a thermometer in an instrument shelter about 2 m above the ground is called as surface air temperature Surface air temperature (SAT) is one of the most important variables

in weather and climate studies, and its diurnal cycle is also needed for a variety of applications The true daily mean obtained from a thermograph is approximated by the mean of 24 hourly readings and may differ by 1.0 degrees C from the average based

on minimum and maximum readings

The surface air temperature database has been extensively reviewed on several earlier occasions, most notably by Wigley et al (1985, 1986) and Ellsaesser et al [1986] and by the Intergovernmental Panel on Climate Change (IPCC) (Folland et al.,

1990, 1992; Nicholls et al., 1996) This work extends these by using slight improvements to the basic surface database and updates the records to the end of 1998 Most importantly, it also provides a comprehensive analysis of surface air temperatures in absolute rather than anomaly units, by the development of a 1 by 1 grid box climatology for all land, ocean, and sea ice areas based on the 1961-1990 period

To document air surface temperature, scientist’s measure land surface temperatures a short distance above 2 meters of the ground at stations around the world These researchers standardize the measurements by accounting for elevation, latitude, time of observation, and type of instrument, and then integrate the information to form a long-term record at a particular location Satellite remote sensing is one of the most commonly used techniques for SAT retrieval One important SAT datasets retrieved by satellite remote sensing technique is available

from AIRS instruments of Aqua satellite AIRS had demonstrated exceptional accuracy and yield, providing over 300,000 temperature and water vapor profiles globally with accuracies comparable to radiosondes-even from uninhabited or totally inaccessible areas

SAT data is also collected from ground-based devices; the SAT data from reanalysis model is the product of numerical weather prediction models Reanalysis are retrospective numerical weather prediction (NWP) with a model with assimilation

of atmospheric observation from radiosonde, satellites, aircraft etc at hence widely use (Wang at 2015) Till now, there are three generations of reanalyzed datasets The third generation reanalysis datasets are better quality in cooperation to older two generation

of reanalysis (Trenberth et al 2011; Schroder et al 2013) One of the third generation reanalysis data have been use on this paper was European Center for Medium-range Weather Forecasts (ECMWF) In this study, data from AIRS satellite and ECWM model will be used as the two main sources for data collection for SAT

Located in the continent of Asia, Vietnam covers 310,070 square kilometers of land and 21,140 square kilometers of water, making it the 66th largest nation in the world with a total area of 331,210 square kilometers Vietnam`s Climate can be divided a tropical and a temperate zone It is characterized by strong monsoon influences, has a considerable amount of sun, a high rate of rainfall, and high humidity that makes it sometimes feel quite uncomfortable Regions located near the tropics and

in the mountainous regions have a slightly cooler, more temperate climate The annual average temperature ranges from 22°C to 27°C year-round There are almost no significant differences in temperature in the southern parts of Vietnam, while the northern regions can be quite cold in the winter There are essentially four distinct

seasons, which are most evident in the northern provinces There are two distinguishable seasons in the southern areas The cold season occurs from November

to April and the hot season from May to October The northern parts of Vietnam have essentially four distinct seasons, it can be quite cool in the winter there, but very warm

in summer Based on the differences in climate and topography between regions, the

research ‘‘ A comparative study of surface air temperature between satellite and model

1.2 Research’s objectives

The objective of this study was to compare SAT data from AIRS and ECWM This study used parameters of correlation coefficient (CC), average different and root mean square error (rmse) to evaluate the SAT and find out which instrument

is better use for SAT

1.3 Research question

- What is the difference between satellites and model data?

- Which dataset is better to use for SAT?

- Are the figures showing climate change in Vietnam?

1.4 Limitations

Because the time for an internship was too short, this research project cannot perform any other calculation method to compare the surface air temperature

PART II LITERATURE REVIEW 2.1 Surface air temperature

2.1.1 Surface air temperature and factors affecting SAT

Surface air temperatures are both an indicator and a driver of regional and global changes There are year-to-year and regional differences in air temperatures due

to natural random variability The global rate of temperature increase has slowed in the last decade (Kosaka and Xie, 2013) Besides, the temperature of a surface is

determined by the balance among the various energy flows that move across it Net radiation—the balance between incoming shortwave radiations and outgoing longwave radiation—produces a radiant energy flow that can heat or cool a surface

Five important factors influence air temperature (Cikgu Geography, 2010):

- Latitude Daily and annual cycles of insolation vary systematically with latitude, causing air temperatures and air temperature cycles to vary as well Yearly insolation decreases toward the poles, so less energy is available to heat the air But because the seasonal cycle of insolation becomes stronger with latitude, high latitudes experience a much greater range in air temperatures through the year

- Surface type Urban air temperatures are generally higher than rural temperatures City surface materials—asphalt, roofing shingles, stone, brick—hold little water, compared to the moist soil surfaces of rural areas and forests, so there is little cooling through evaporation Urban materials are also darker and absorb a greater portion of the Sun’s energy than vegetation-covered surfaces The same is true for areas of barren or rocky soil surfaces, such as those of deserts

- Coastal or interior location Locations near the ocean experience a narrower range of air temperatures than locations in continental interiors Because water heats

and cools more slowly than land, air temperatures over water are less extreme than temperatures over land When air flows from water to land, a coastal location will feel the influence of the adjacent water

- Elevation Temperature decreases with elevation At high elevation, there are fewer atmospheres above the surface, and greenhouse gases provide a less effective insulating blanket More surface heat is lost to space On high peaks, snow accumulates and remains longer The reduced greenhouse effect also results in greater daily temperature variation

- Atmospheric and oceanic circulations Local temperatures can rise or fall rapidly when air from one region is brought into another Temperatures of coastal regions can be influenced by warm or cold coastal currents

2.1.2 The important of Surface air temperature

Air temperature conditions many aspects of human life, from the clothing we wear to the fuel costs we pay Air temperature and air temperature cycles also act to select the plants and animals that make up the biological landscape of a region (Stisen et al., 2007; Shamir et al., 2014; Benali et al.,2012) And air temperature, along with precipitation, is a key determiner of climate, which will explore in more depth in this research SAT is an important parameter for a wide range of applications such as vector-borne disease bionomics, hydrology and climate change studies Air temperature data is usually obtained from measurements made in meteorological stations, providing only limited information about spatial patterns over wide areas The use of remote sensing data can help overcome this problem, particularly in areas with low station density, having the potential to improve the estimation of SAT at both

regional and global scales Some studies have tried to derive maximum (Tmax), minimum (Tmin) and average air temperature (Tavg) using different methods

2.1.3 Global Surface air temperature change

The Intergovernmental Panel on Climate Change has accepted the invitation to generate a special report on the impacts of global warming of 1.5 C by 2018 As part

of an ambitious and urgent plan, a wide-ranging analysis on the effects of limiting global warming to 1.5 C is required Various studies have been conducted regarding the 2.0 C target (Joshi et al., 2011; Jiang and Fu, 2012; Zhang, 2012;Zhang et al., 2013; V1autard et al., 2014; Sui et al., 2015;Jiang et al., 2016) For instance, by analyzing the result simulated by Coupled Model Inter comparison Project Phase 3(CMIP3) models, Jiang and Fu (2012) pointed out that the 2 C global warming is projected to occur in 2064, 2046 and 2049 under the SRES B1, A1B and A2 in multi-model ensemble (MME), respectively, but would occur in a 30-40 year period in the individual models For the Representative Concentration Pathways (RCPs), the globally averaged temperature would cross 2 C threshold sometime between2030 and

2060 under RCP4.5 and RCP8.5 (Zhang, 2012; Vautard et al., 2014; Sui et al., 2015; Jiang et al., 2016)

Regional climates are the result of complex processes that vary strongly with location and therefore respond differently to changes in global-scale influences For instance, the CMIP3 MME suggested that the annual mean surface air temperature (SAT) averaged over China would be elevated by 2.7-2.9 C when the 2 C target occurred, but the scope of warming is quite large (approximately 2.3-3 C) among the individual models (Jiang and Fu, 2012) These authors also showed that the warming

is stronger towards the north and on the Qinghai-Tibetan Plateau Based on the simulations of the Coupled Model Inter comparison Project Phase 5 (CMIP5) models, the MME showed that the annual mean SAT averaged over China would increase by 2.6 C under a 2 C global warming (Suiet al., 2015), and Northwest China shows the fastest warming trend and followed by Central North China and Northeast China (Zhang, 2012)

Unlike the 2 C target, the 1.5 C target is relatively unexplored, and there is a relative lack of scientific knowledge about the implications of 1.5 C warming (Hulme, 2016;Mitchell et al., 2016; Schleussner et al., 2016; Zhai et al.,2017; Zhang et al.,

2017 ) The CMIP5 MMEs showed that the global mean 1.5 C warmings related to the pre-industrial level (1861e1900) would occur by approximately 2036, 2028,2033 and 2025 under RCP2.6, RCP4.5, RCP6.0 and RCP8.5,respectively (Hu et al., 2017)

At the moment, the mean temperature over Asia would increase by 2.3 C, with stronger warming in high latitudes than in low latitudes, and it would be 0.5-1 C) lower than that under the 2 C target (Xuet al., 2017) A warming of 20.5 C (from 1.5-

2 C) would lead to significant increases in the temperature and precipitation extremes

in most regions (Wang et al., 2017)

2.1.4 Climate change in Vietnam

One of the biggest manifestations of the change in air surface is climate change According to the Intergovernmental Panel on Climate Change (IPCC) in the Fourth Assessment Report (IPCC, 2007) climate change is a change in the state of the climate system that can be received Knowing the variation in the average and the volatility of its attributes is maintained for a long enough period, typically decades or longer

Climate change research in Vietnam has been conducted since the 1990s before (Nguyen Duc Ngu, 1991,1999; Nguyen Trong Hieu, 1993) Over the past 50 years, the annual average temperature has risen by about 0.5ºC in the whole country (Ministry of Natural Resources and Environment, 2009) According to the IPCC (2007), during the period 1906-2005 the average global air temperature increased 0.74 ± 0.18 ° C The years 2005 and 1998 were the hottest years since 1850 In the 12 years from 1995-

2006, there were 11 years of the hottest years since 1850, except for 1996 The number

of days of wintering decreases in most areas On average, the number of extremely hot days (10% of the hottest days) increased and the number of extremely cold days (10%

of the coldest days) decreased Heavy rainfall events have increased in many parts of the continent since around 1950, even in places where total rainfall has declined It has been observed that rare rainfall is recorded (once in 50 years) Longer, longer-lasting droughts have been observed in many areas with wider ranges, particularly in the tropics and subtropics since the 1970s High temperatures and precipitation on continental areas is one of the causes of this phenomenon Increased temperatures affect public health, as the number of deaths increases due to hot waves

Climate change is expected to worsen the already worrisome impacts of natural disasters on the population of Viet Nam, which are felt primarily by the poor and vulnerable, especially women, children, and the elderly Four main sources of health impacts are extreme weather events causing injuries and deaths, water contamination,

infectious diseases, food shortages, and mental health problems; droughts and heavy

rainfall causing significant reductions in crop yields and productivity of subsistence agriculture, which may lead to malnutrition, micronutrient deficiencies, or, in more extreme cases, starvation an increase in the number of very hot days in large cities,

along with forest fires and dust storms adversely impacting air quality over broad areas (both urban and rural) and exacerbating the occurrence and intensity of health complications associated with high temperatures (e.g., heat stroke) and respiratory diseases (e.g., asthma); and changes in temperature and rainfall patterns impacting not only the occurrence of vector-borne diseases such as malaria and dengue, but also changing and possibly extending the geographical habitat of the vectors of such

on density profiles of atmospheric constituent gases responsible for molecular absorption, there is a weighting function for each of the 2378 AIRS channels (with wavelengths between 3.7 and 15.4 μm) By measuring the infrared radiance at each of these AIRS channels and calculating temperature using the Planck equation, AIRS provides the data necessary for the calculation of vertical temperature profiles from the surface (or top of thick cloud cover) to more than 70 km in height (JPL, 2014)

The Atmospheric Infrared Sounder (AIRS) is a cross-track rotary scanning sensor with a 49.5 degrees (from nadir) ground coverage as well as views to cold space and onboard spectral and radiometric calibration sources per scan cycle Its coverage is

pole-to-pole and covers the globe two times per day (daytime and nighttime) The scan cycle has a period of 8/3 seconds during which ninety ground footprints are observed Each 22.4 ms footprint contains a spectrum with all 2378 spectral samples The IR spectral regions covered are 649 - 1136, 1216 - 1613, and 2164 - 2674 The IR spatial resolution of AIRS is 13.5 km at nadir from the 705.3 km orbit The swath width is 1650 km AIRS flies on NASA's Aqua satellite with a daytime equator crossing of roughly 13:30 local time on the ascending orbit pass

AIRS data are used by weather forecasting centers around the world By incorporating AIRS measurements into their models, forecasters have been able to extend reliable mid-range weather forecasts by more than six hours (NASA/NOAA Announce Major Weather Forecasting Advancement) AIRS data have also improved forecasts of the location and magnitude of predicted storms AIRS temperature and water vapor profiles are available in real time to regional weather forecasters, providing twice-daily weather measurements for the entire Pacific Ocean, once in the morning and once in the evening In mid-2004 the Joint Center for Satellite Data Assimilation (JCSDA) demonstrated significant impact from AIRS data in both the northern and southern hemispheres (Le Marshall 2005a,b) This was achieved through use of an enhanced spatial and spectral AIRS observational dataset in conjunction with

an analysis methodology that paid additional attention to the possible presence of clouds

2.2.2 ECMWF model

The European Centre for Medium-Range Weather Forecasts (ECMWF) is an independent intergovernmental organisation supported by most of the nations of Europe and is based at Shinfield Park, Reading, United Kingdom It operates one of

the largest supercomputer complexes in Europe and the world's largest archive of numerical weather prediction data

ECMWF was established in 1975, in recognition of the need to pool the scientific and technical resources of Europe's meteorological services and institutions for the production of weather forecasts for medium-range timescales (up to approximately two weeks) and of the economic and social benefits expected from it 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

ECMWF makes significant contributions to support research on climate variability, pioneering an approach known as reanalysis This involves feeding weather observations collected over decades into a NWP system to recreate past atmospheric, sea- and land-surface conditions over specific time periods to obtain a clearer picture

of how the climate has changed Reanalysis provides a four-dimensional picture of the atmosphere and effectively allows monitoring of the variability and change of global climate, thereby contributing also to the understanding and attribution of climate change (Sky News, 2015)

2.2.3 Matlab software

The name MATLAB stands for MATrix LABoratory is a multi-paradigm numerical computing environment and proprietary "programming" language developed by MathWorks (The MathWorks Inc., 2005) MATLAB allows matrix manipulations, plotting of functions and data, implementation of algorithms, creation

of user interfaces, and interfacing with programs written in other languages, including

C, C++, C#, Java, Fortran and Python

MATLAB is a high-performance language for technical computing It integrates computation, visualization, and programming environment Furthermore, MATLAB is a modern programming language environment: it has sophisticated data structures, contains built-in editing and debugging tools, and supports object-oriented programming These factors make MATLAB an excellent tool for teaching and research

MATLAB consists of 5 main parts:

- Development Environment: A set of tools that allows us to use MATLAB functions and files It includes: desktop matlab, command window, a command history,

an editor, debugger, browsers for viewing help, the workspace, files, the path search

- MATLAB Mathematical Function Library: A set of mathematical functions such as sum, sine, arithmetic, and so on

- MATLAB Language (script): high-level programming language

- Graphics: tools that help visualize data in graphs In addition, it allows to build graphical interface

- MATLAB Application Program Interface (API): The library allows you to use the MATLAB calculation functions in C or FORTRAN

It has powerful built-in routines that enable a very wide variety of computations It also has easy to use graphics commands that make the visualization of results immediately available Specific applications are collected in packages referred

to as toolbox There are toolboxes for signal processing, symbolic computation, control theory, simulation, optimization, and several other fields of applied science and engineering

PART III METHODOLOGY

- ECMWM data: ERA Interim, Synoptic Monthly Means, with variable: 2 meter temperature (t2m) The data is available at

http://apps.ecmwf.int/datasets/data/interim-full-daily

Collecting data

(Model data and Satellite data)

Data analysis

( Matlab and Excel)

Comparison of satellite and model data (CC, Average difference and rmse)

Both of 2 dataset with spatial resolution is 1 x 1 and monthly scale from 2003

to 2014 The range of region for data use was shown in Fig 2, 3, 4 and 5 The downloaded data has been saved as ‘’dataname.hdf’’ file

3.2 Region selected

The research area will be of interest to the SAT data of Vietnam and three sub regions are Bac Bo (the north), Tay Nguyen (High land) and Nam Bo (the south)

The region selected:

Fig 3 Bac Bo region with latitude interval - ′ and longitude interval

′ - ′ (from Google earth)

Fig 5 Nam Bo region with latitude interval ′ - ′ and longitude

3.3 Data Analysis

- Step 1: Read and extract data

After the data is downloaded and saved as ‘’dataname.hdf’’ file Matlab software was used to read the data with the command:

1 filename=’’source where the file is stored’’: to run the hdf file

2 ncdisp(filename): to read the content of the file

3 variable name=ncread(‘’source where the file is stored’’,’’variable name’’):

to read values in variable which ECMWF model is t2m and AIRS is SurfSkinTemp_A and SurfAirTemp_D

Then once the data has been read, the monthly SAT data has been extract to Excel file and saved as ‘’dataname.xls’’ The data of each month recorded corresponds

to the selected data area

- Step 2: Aggregation of data

In term of ECMWF model data, corresponding to the 10 years selected 120 monthly SAT tables of study region and 360 tables of the sub regions are averaged by using average function of Excel

In term of AIRS satellite data, day and night SAT data for 1 month was aggregated and calculated as the mean of one month because the ECMWF model was not divided into day and night data Similarly, 120 monthly SAT tables of study region and also 360 tables of the sub regions are averaged by using average function of Excel

- Step 3: Compare data

By using correlation coefficient, average different and root mean square error to compare and evaluate SAT data between the AIRS satellite and the ECMWF model:

!! = ∑ (%+ & − %̅)(*& − *̅)

&,-.∑ (%+ &− %̅)

- %& and *& are the surface air temperature estimated by AIRS satellite and ECMWF model respectively

- %̅ and *̅ are their arithmetic mean

- n: number of months considered

PART IV RESULTS 4.1 The different between AIRS satellite and ECMWF model about SAT

4.1.1 The different between AIRS satellite and ECMWF model about SAT in sub regions

The vast delta area is located in the Red River Basin, covering an area of 14.8 thousand km2 and 4.5% of the national area The delta is in triangular form, the top is Viet Tri City and the bottom is the east coast line This is the second largest delta in Vietnam (after the Mekong Delta area of 40,000 km2) by the Red River and the Thai Binh River Much of the plain surface is quite flat, with a height of 0.4 - 12m above sea level

All year round, the temperature is relatively high and humid; the climate is influenced by mainland China and continental climate While part of the coastal zone

is affected by subtropical climate and wet monsoon from the mainland

The region has a tropical monsoon climate year-round with two distinct summer and winter seasons At the same time, the annual monsoons are affected by the northeast monsoon and southeast monsoon The annual average temperature increases

from the north to the south and has a climate of harmony, which is characteristic of the Northern Delta and coastal areas Summer weather from April to October is hot and humid and rainy until the monsoon The winter is from November to March with dry, drizzly rain The annual average temperature is about 25 degrees Celsius, the average rainfall is between 1,700 and 2,400 mm In winter the temperature drops to the lowest

in December and January This time, in the northern mountainous areas (such as Sa Pa, Tam Dao, Hoang Lien Son), the temperature remains below 0 degrees Celsius, appearing frosty and likely snowfall

The climate in Bac Bo is often affected by bad weather, with an annual average

of 6 to 10 storms and tropical low pressure causing floods, direct threats to life and agriculture of the whole local area

- AIRS satellite an ECMWF model data of surface air temperature of Bac bo region (◦C)

Fig 6 AIRS measured and ECMWF model estimated over Bac bo region

In the Bac bo area, the SAT of AIRS range from 15-25 degrees, while the SAT

of the ECMWF model had a greater amplitude range of 8-22 degrees This sub region has highest average different -4.3◦C with CC was the same with study region Vietnam;

Year

rmse was also accounted for the highest number 4.81◦C Besides, the data from the hot season and the cold season did not have similarities with the CC of hot season was 0.81◦C and cold season was 0.82◦C (Table 12) Besides, the coldest months in both dataset are in November, December, January and February and the hottest months are July, August and September, the monthly bias has shown the different between that month of 2 datasets (in appendix) Overall, with the larger amplitude and lower annual temperature in compare with the annual temperature of Bac Bo region (Table 4), ECMWF model seem to be less accurate than AIRS satellite data

4.1.1.2 Tay nguyen region

- General information

Tay Nguyen is bordered by Quang Nam Province to the North, Quang Ngai, Binh Dinh, Phu Yen, Khanh Hoa, Ninh Thuan and Binh Thuan to the East, Dong Nai, Binh Phuoc, bordered by Attapeu (Laos) and Ratanakiri and Mondulkiri (Cambodia) provinces While Kon Tum has its western border with both Laos and Cambodia, Gia Lai, Dak Lak and Dak Nong share the border with Cambodia Lam Dong has no international border Considering the Central Highlands as the total area of 5 provinces, the Central Highlands is 54,641 2 / (General Statistics Office of Vietnam, 2012)

In essence, Tay Nguyen is not a unique plateau but a series of adjoining plateaus Kon Tum Highlands, Kon Ha Nung, Pleiku is about 800 m high, M'Drak is about 500 m high, Buon Ma Thuot is about 500 m high, Mo Nong is about 800-1000

m high, Lam Vien is about 1500 m high and Di Linh is about 900-1000 m high All of

these plateaus are bounded to the east by the mountain ranges and the main highland is the Southern Truong Son

Tay Nguyen can be divided into three terrain sub regions and three climates, including North Central Highlands (corresponding to Kon Tum and Gia Lai), Central Highlands (corresponding to Dak Lak and Dak Nong), South Highlands (corresponding to Lam Dong province) Central Highlands have lower elevations and higher temperatures than the two northern and southern sub-regions

sub-Located in the tropical zone, the climate in Tay Nguyen is divided into two seasons: the rainy season from May to October and the dry season from November to April, with March and April being the two warmest and driest months Due to the influence of the altitude, while in the highlands of 400-500 m high, the climate is relatively cool and rainy, with the high plateau of over 1000 m; the climate is cool all year round, characterized by high mountain climate

- AIRS satellite an ECMWF model data of surface air temperature of Tay nguyen region (◦C)