Research Article Evaluation of Different Soil Salinity ...downloads.hindawi.com/journals/js/2016/7596175.pdfResearch Article Evaluation of Different Soil Salinity Mapping Using Remote

Research ArticleEvaluation of Different Soil Salinity Mapping Using RemoteSensing Techniques in Arid Ecosystems Saudi Arabia

Mohamed Elhag

Department of Hydrology and Water Resources Management Faculty of Meteorology Environment amp Arid Land AgricultureKing Abdulaziz University Jeddah 21589 Saudi Arabia

Correspondence should be addressed to Mohamed Elhag melhagkauedusa

Received 14 July 2015 Revised 12 November 2015 Accepted 2 December 2015

Academic Editor Guiyun Tian

Copyright copy 2016 Mohamed Elhag This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Land covers in Saudi Arabia are generally described as salty soils with sand dunes and sand sheets Waterlogging and highersoil salinity are major challenges to sustaining agricultural practices in Saudi Arabia principally within closed drainage basinsAgricultural practices in Saudi Arabia were flourishing in the last two decadesThe newly reclaimed lands were added annually anddistributed all over the country Irrigation techniques are mostly modernized to fulfill water saving strategies Nevertheless waterresources in SaudiArabia are under stress and groundwater levels are depleted rapidly due to heavy abstraction thatmay exceed cropwater requirements inmost of the cases due to high evaporation ratesThe excess use of irrigational water leads to severe soil salinityproblems Applications of remote sensing technique in agricultural practices became widely distinctive and cover multidisciplinaryprincipal interests on both local and regional levels The most important remote sensing applications in agricultural practices arevegetation indices which are related to vegetation and water especially in an arid environment Soil salinity mapping in an aridecosystem using remote sensing data is a demanding task Several soil salinity indices were implemented and evaluated to detectsoil salinity effectively and quantitatively Thematic maps of soil salinity were satisfactorily produced and assessed

1 Introduction

Saudi Arabia has practiced an eminent development in vitalsectors over the last three decades As a result rapid escalationin terms of water demands in domestic industrial andagricultural activities have been countersigned Agricultureis the keystone water consumption sector it consumes about85 of total national water use (Ministry of Water and Elec-tricity) [1] During the period of 1974ndash2006 Saudi Arabiangovernment subsidized the agricultural sector to improveliving standards in urban areas and to achieve sustainabilityThese subsidies lead to significant increases in agriculturalareas primarily for rigorous irrigated crops such as wheatmaize and alfalfa the irrigated areas increased from less than400000 ha in 1971 to about 162 million ha in 1992 [2 3]The extensive groundwater extrication subsequently leads tohigher soil salinity rates in most of the agricultural zones [4ndash6]

Excessive irrigation of less developed drained soil hasled to high floating water tables waterlogging and soil

salinization [7] In the absence of well drainage network inshallow soil water table depth intends to develop remarkablyand eventually irrigable lands will be abandoned [8] Inthe incidence of a higher soil salinity crop production canbe dramatically affected when salts accumulate in the soilsurface [9ndash11] Conservatively soil salinity and waterloggingcan be regulated through sustaining the salt concentrationat the root zone below the crops constraint [12] This canbe realized by maintaining the equilibrium between soilrecharge and discharge by draining the surplus irrigationalwater [13]

Dwivedi [14] used the first four bands ofMSS to designatesalty soils Chaturvedi et al [15] and Singh and Srivastav [16]used brightness and thermal temperature layers to detect soilsalinity Menenti et al [17] used 3 bands of TM (1 5 and7) to detect salt minerals Moreover Mulders and Epema[18] used three different bands of TM (3 4 and 5) to detectcalcareous surfaces as an indication of high salt content Sahaet al [19] used the thermal infrared band of TM to classifyinland waterlogged areas Rao et al [20] used the false color

Hindawi Publishing CorporationJournal of SensorsVolume 2016 Article ID 7596175 8 pageshttpdxdoiorg10115520167596175

2 Journal of Sensors

composition of TMbands (2 3 and 4) to distinguish betweenmoderate and high soil content soils Naseri [21] used bothunsupervised and supervised classification algorithms todelineate soil salinity on a regional scale using three bandsof MSS and four bands of TM Darvishsefat et al [22] usedprincipal component analysis to map soil salinity using fourbands of ETM+ (3 4 5 and 7) Zhang et al [23] used differentbands of OLI-8 to develop a soil salinity detection algorithmbased on the Soil Salinity Information Extraction (SSIE)

Implementation of remote sensing data in soil salinitymapping is firm to employ since salinization is not a staticprocessTherefore a number of derivatives and alternatives tonormalized difference vegetation index have been anticipatedto discourse soils salinization monitoring and mapping [24ndash26] Different remote sensing indices such as the salinityindex (SI) the normalized difference salinity index (NDSI)and the brightness index (BI) in addition to normalizeddifference vegetation index (NDVI) were implemented toinvestigate the means by which these indices work for soilsalinity mapping in the arid environment [27 28] Thesalinity index (SI) which combines the blue and red bands issensitive to the surface reflectance of salt-affected land withsparse vegetation cover [27]

The perspective of using remote sensing data and GISpractices has been demonstrated in several scholarly worksto be the most resourceful [29ndash32] The comprehensive useof satellite remote sensing and GIS has been recognized tobe a cost-effective method for monitoring soils salinizationin poorly drained basins [33 34] However state-of-the-artoutcome specifies that there is no ideal combination of datatype and questioning method which can be functional withidentical realization across approximately several environ-mental conditions [25ndash27]

The aim of current research project is to differentiatethe salinized soils from nonsalinized soils implementingdifferent approaches of digital image classification and bandmath techniques To reach that objective several soil salinityindices were applied to the Landsat 8 data verified withground truth data to validate the percentage of estimationsaccuracy based on different errorrsquos assessment approach

2 Materials and Methods

21 Study Area The study area Wadi Al Dawasir townis located in the plateau of Najd at latitude of 44∘431015840 andlongitude of 20∘291015840 about 300 km south of the capital cityRiyadh (Figure 1) This study area is comprised of gravellytableland disconnected by insignificant sandy oases andisolated mountain bundles Across the Arabian Peninsulaas a whole the tableland slopes toward the east from anelevation of 1360 meters in the west to 750 meters at itseasternmost limit Wadi Al Dawasir and Najran regions arethemajor irrigationwater abstraction fromAl-Wajid AquiferAgriculture in Wadi Al Dawasir area consists of technicallyhighly developed farm enterprises that operate modern pivotirrigation system The size of center pivot ranges from 30 hato 60 ha with farms managing hundreds of pivots withthe corresponding number of wells The main crop grown

Figure 1 Location of the study area

in winter is wheat and occasionally potatoes tomatoes ormelons All year fodder consists of alfalfa which is cut up to10 times a year for food Typical summer crops for fodder aresorghum and Rhodes grass which is perennial but dormantin winter The shallow alluvial aquifers could not sustainthe high groundwater abstraction rates for a long time andgroundwater level declined dramatically in most areas Mete-orological features of the area are speckled Five elements ofmeteorology are constantly recorded through fixed weatherstation located within the study area Temperature variesfrom 6∘C as minimum temperature to 43∘C as maximumtemperature Relative humidity is mostly stable at 24 Solarradiation of average sunrise duration is generally 11 hrsdayAverage wind speed is closer to 13 kmhr and may reach upto 46 kmhr in thunderstorm incidents Finally mean annualrainfall is about 376mm [5]

22Methodological Framework The current research work isbased on assessing a regression correlation between groundtruth data and its corresponding salinity indexrsquos valuesconducted from satellite images Therefore accurate syn-chronization of ground truth data collection and satellitebypassing were exercised

221 Soil Sampling and Analysis A total number of 150random soil samples were collected from Wadi Al Dawasirarea with a minimum distance of 100m between the sampleslocations to avoid data clumping (Figure 2) The standardprocedure of determining soil extract salinity in terms ofElectrical Conductivity (EC) was followed according to Shaw[35] under laboratory condition

222 Remote Sensing Data Enhancement Remote sensingdata was obtained from Landsat Operational Land Imager(OLI-8) which was acquired on April 17th 2014 Typi-cal atmospheric and radiometric corrections and spatialresolution enhancement were implemented for each bandindividually Furthermore overcorrection was analyzed forthe red and near infrared bands [36] Atmospheric correctionis basically considering blue band offset Therefore thelargest atmospheric consequence is theoretical to be the mosttruthful value within the blue band The calibrated radiancevalue of this offset allows deciding 120581 valueTherefore 120582minus120581 rule

Journal of Sensors 3

0 125 25 50

(km)

N

WS

E

Figure 2 Soil sample location on false-color Landsat 8 image

the offset values for the green red and near infrared bandunder investigation [37]

223 Soil Salinity Indices Soil salinity detection usingremote sensing techniques developed through several schol-arly works LandsatMultispectral Scanning (MSS) TM (The-matic Mapper) and EnhancedThematic Mapper (ETM+) inaddition to Operational Land Imager (OLI-8) images werethe most used remote sensing data in soil salinity mappingSeveral algorithms of detecting and mapping soil salinitywere developed because of the sensors andor study areavariability There were many attempts to use the remotesensing data and its techniques to detect the salt-affectedsoil Soil salinity indices are principally adjusted to detectsalt mineral in soils based on the different responses of saltysoils to various spectral bands Table 1 summarizes differentindices used to map soil salinity in current study

23 Validation Linear regression model was used to vali-date the correlation established between the ground truthsdata measured from the 150 soil samples collected and thecorresponding salinity indexrsquos values estimated using theaforementioned equations Root square root square adjustedroot mean square error and mean of response are used tosignify the inequality of variance and correlation of the linearregression model [38]

3 Results and Discussion

Theuse of different soil salinity equations gives the results thedifferential preferences to achieve an adequate soil salinityestimation on a large scale using remote sensing data Theevaluation of different soil salinity indices was based on theregression model fit effect significance and model parame-ters Table 2 summarizes different regression significances

y = 25315x + 18571

10

20

30

40

0 2 4 6 8

Estim

ated

soil

salin

ity v

alue

s (pp

m)

Thou

sand

s

Measured soil salinity values (ppm) Thousands

R2 = 09764

Figure 3 Linear regression between measured and estimated soilsalinity values

Based on different ldquoerrorsrdquo estimated by analysis ofvariance SI-9 represents the most straightforward linearregressionmodel Figure 3 shows the correlation between theestimated soil salinity values using soil salinity index-9 andthe collected ground truth values

Due to the fact that the soil salinity indices used inthe current research are varied (11 in total) categorizationof different indices using principal component analysis willhelp decide which is similardifferent to which Figure 4represents the grouping to different indices according toprincipal component analysis on covariances

Generally different soil salinity indices fell into twogroups SI-6 SI-9 SI-3 and BI are in one group and the restare in the other group SI-6 and NDSI failed to join theircorresponding group and remain as group outlier [39] Lackof correlation is the main reason of the insignificance of SI-6and NDSI High correlation significance between SI-3 SI-9and BI is the best to represent soil salinity indices accordingto Table 2

The results proved the acceleration in the rate of vegeta-tion death and soil salinization based on the high correlationbetween higher soil salinity values and negative normalizeddifference vegetation index with 1198772 values of 092 [40]Moreover higher soil salinity was related to the improper andintense drainage systems set up in addition to the lack of anoperative water resource management plan [41]

The dynamicity of the soil salinization process addedfurther complications to designating salt-affected soils in asystematic uniform perspective The use of different algo-rithms based on implementing different combinations andorratios of OLI-8 bands in the form of soil salinity indicesevidenced to be more efficient to overcome soil dynamicityproblems [23 42]

The selection of the sensitive bands adequate for propersoil salinity mapping is not systematically generalized [4247] Spatial variability and land cover variations are themost limiting factors of band selection of a sensor [23 48]


Table 1 Soil salinity indices based on different band ratios of Landsat

Salinity index Band ratios Reference

Normalized difference salinity index NDSI = (119877 minus NIR)(119877 + NIR)

[43]

Brightness index BI = radic(1198772 + NIR2) [44]

Salinity index-1 SI = radic(119861 times 119877) [44]Salinity index-2 SI = radic(119866 times 119877) [44]

Salinity index-3 SI = radic(1198662 + 1198772 + NIR2) [27]

Salinity index-4 SI = radic(1198662 + 1198772) [27]

Salinity index-5 SI = 119861119877

[45]

Salinity index-6 SI = (119861 minus 119877)(119861 + 119877)

[45]

Salinity index-7 SI = (119866 times 119877)119861

[45]


[46]

Salinity index-9 SI = (NIR times 119877)119866

[46]

119861 blue band 119866 green band 119877 red band NIR near infrared band of Landsat 8 image

Table 2 Regression significances

NDSI BI SI-1 SI-2 SI-3 SI-4 SI-5 SI-6 SI-7 SI-8 SI-9119877 square 062 094 089 090 095 090 044 044 091 086 098119877 square adjusted 062 094 089 090 095 090 044 044 091 086 098Root mean square error 151 76008 100038 113967 77306 163887 029 029 178057 126635 60314Mean of response 001 30262 14749 15716 33634 22431 018 001 19408 14958 28103

100500minus05minus10

10

05

00

minus05

minus10

Component 1 (976)

Figure 4 Variance based principal component analysis

Therefore some of the used equationsmay have better resultsin specific areas implementing specific band ratios than

others taking into consideration the use of OLI-8 as a sourceof remote sensing data [23 49]

Estimated values of soil salinity index-9 were then usedto map and classify soil salinity in the whole study areaRegular interpolation technique with six different smoothingKernel functions is implemented and evaluated according toKrivoruchko et al [50] as shown in Table 3 and Figure 5

In Figure 5 only 3 of Wadi Al Dawasir area is catego-rized to be with low soil salinity followed by 10 of moderatesoil salinity classMeanwhile more than half of the study areais subjected to a higher concentration of soil salinity (52)In addition there are soils in Wadi Al Dawasir which cannotbe used for the current agricultural practice due to extremecondition of soil salinity (35)

Kernel function of polynomial-5 was used to produce athematic soil salinity map in the study area (Figure 6) Basedon Soil Survey approach in conjunction with natural breaksclassification method was used to reclassify the predictionrisk maps into four classes according to their toxicity levelsinto Low Moderate High and Extreme [51 52] The role ofthe classifier is based on squared error minimization from aclassrsquos means by creating internally homogenous groups butmaintaining heterogeneity between classes

Figure 7 shows the percentages of Wadi Al Dawasiragricultural area which were fallen into the four soil salinityclasses based on NDVI estimationThe total agricultural area


Table 3 Evaluation of different interpolation kernels function

Kernel functionExponential Gaussian Quartic Epanechnikov Polynomial-5 Constant

Regression function 0006119883 + 29339 minus0001119883 + 29659 0003119883 + 29477 minus0005119883 + 29778 0005 + 29368 minus0001119883 + 29679

Mean 1383 1421 1378 1412 1355 1428Root mean square 4251 4257 4250 4257 4253 4266Mean standardized 00065 00067 00065 00066 00064 00067Root mean square standardized 00201 00201 00202 00201 00202 00200Average standard error 211114 211797 210334 211526 209641 212674

10

3552

3

LowModerate

HighExtreme

Figure 5 Soil Salinity classes in Wadi Al Dawasir area

0 25 50

(km)Soil salinity

LowModerate

HighExtreme

N

W

S

E

Figure 6 Soil salinity map in Wadi Al Dawasir

0

10

20

30

40

50

Low Moderate High Extreme

Tota

l agr

icul

tura

l are

a (

)

Soil salinity class

Figure 7 Contribution of different soil salinity classes to the totalagricultural area

in Wadi Al Dawasir expressed low soil salinity contribution(65) Those areas are the best to preserve the naturalresources and the agricultural practice in Wadi Al DawasirAbout 36 of the total agricultural area constantly needsefforts to sustain the current agriculture practice Both of theaforementioned classes are less than the half of the croppedarea of Wadi Al Dawasir Closer to 58 of the cultivated areais on improper soils due to its high salinity content

Soil salinity index-9 as a remotely sensed indicator typi-cally depends on the vegetation growing conditionswhich arehelpful for mapping spatially the distribution of soil salinity[53] The response of certain features of crop stress dehy-dration and senescence using vegetation indices have beenexamined to determine their response in different spectra[54 55] According to Carter [54] cell structure damageand chlorophyll reduction could be reliably measured usingreduced near infrared reflectance (NIR) and increased visiblereflectance (VIS) among several plant species in response tostress Tilley et al [53] also investigated salt stress responses inboth of VIS andNIR Normalized difference vegetation index(NDVI) as one of the most implemented vegetation indiceshas been inspected in abundant scholarly work by using thedifferent vegetation reflectance to estimate soil salinity [40]

4 Conclusions

Crop production in Wadi Al Dawasir is completelydependent on underground water resources represented in


artesian wells The use of the underground water resourcesis annually overexploited due to the horizontal expansion ofthe agricultural land The driving forces of the agriculturalland expansion and the use of the underground waterresources are controlled only by the Ministry of Water andElectricity but not by the local inhabitants Such uncontrolledconditions put the water resources and the soils of Wadi AlDawasir in jeopardy The use of excess irrigational waterto cope with the extremely arid conditions in the studyarea continuously creates more drainage problems as theagricultural areas in Wadi Al Dawasir have no drainagesystem by any means Water logging and soil salinizationare common problems in agricultural areas describedby no drainage system Consequently the hydrologicalsceneries and the geomorphological patterns of arid landsare very important aspects to be considered in any rationaldevelopment plan Soil salinity indices are siteconditionspecific and soil salinity index-9 is the best to representsoil salinity values in the current study area The currentagricultural practices in Wadi Al Dawasir are obviouslynot the optimum water andor soil resources managementMore than half of the study area is characterized by eitherhigh or extreme soil salinity classes Crop reallocation andrational water use may serve the goal of sustainability tokeep the current agricultural practice in Wadi Al DawasirMonitoring and temporal soil salinity investigation are thekeystone exercise required for adequate natural resourcesmanagement plans in the near future

Conflict of Interests

The author declares that there is no conflict of interestsregarding the publication of this paper

Acknowledgment

This project was funded by the Deanship of ScientificResearch (DSR) King Abdulaziz University Jeddah underGrant no 137-155-1436-G The author therefore acknowl-edges and thanks DSR for technical and financial support

References

[1] Ministry of Water and Electricity (MOWE) Annual ReportMinistry of Water and Electricity (MOWE) Riyadh SaudiArabia 2008

[2] A W Abderrahman ldquoWater demand management in SaudiArabiardquo in Water Management in Islam N I Faruqui A KBiswas and M J Bino Eds IDRC 2001

[3] World Bank ldquoA water sector assessment report on the countriesof the co-operationCouncil of theArab States of theGulfrdquo TechRep 32539-MNA World Bank 2005

[4] KHAl-Zahrani ldquoWater demandmanagement in theKingdomof Saudi Arabiardquo Conference of the International Journal of Artsamp Sciences vol 2 no 3 pp 68ndash76 2010

[5] K H Al-Zahrani and M B Baig ldquoWater in the kingdom ofSaudi Arabia sustainable management optionsrdquoThe Journal ofAnimal and Plant Sciences vol 21 no 3 pp 601ndash604 2011

[6] M Elhag and J A Bahrawi ldquoConservational use of remotesensing techniques for a novel rainwater harvesting in aridenvironmentrdquo Environmental Earth Sciences vol 72 no 12 pp4995ndash5005 2014

[7] J M Bradd W A Milne-Home and G Gates ldquoOverview offactors leading to dryland salinity and its potential hazard inNew SouthWales AustraliardquoHydrogeology Journal vol 5 no 1pp 51ndash67 1997

[8] D R Williamson ldquoLand degradation processes and waterquality effects waterlogging and salinityrdquo in Farming ActionCatchment Reaction The Effect of Dry-Land Farming on theNatural Environment J Williams R A Hook and H LGascoigne Eds pp 162ndash190 CSIRO Publishing CollingwoodVa USA 1998

[9] E Houk M Frasier and E Schuck ldquoThe agricultural impactsof irrigation induced waterlogging and soil salinity in theArkansas Basinrdquo Agricultural Water Management vol 85 no1-2 pp 175ndash183 2006

[10] R Jones and G Marshall ldquoLand salinization waterloggingand the agricultural benefits of a surface drainage schemein Benerembah irrigation districtrdquo Review of Marketing andAgricultural Economics vol 60 pp 173ndash189 1992

[11] R George D McFarlane and B Nulsen ldquoSalinity threatens theviability of agriculture and ecosystems in Western AustraliardquoHydrogeology Journal vol 5 no 1 pp 6ndash21 1997

[12] F Konukcu J W Gowing and D A Rose ldquoDry drainage asustainable solution to waterlogging and salinity problems inirrigation areasrdquo Agricultural Water Management vol 83 no1-2 pp 1ndash12 2006

[13] S Abdel-Dayem S Abdel-Gawad and H Fahmy ldquoDrainagein Egypt a story of determination continuity and successrdquoIrrigation and Drainage vol 56 no 1 pp 101ndash111 2007

[14] R S Dwivedi ldquoMonitoring of salt-affected soils of the Indo-Gangetic alluvial plains using principal component analysisrdquoInternational Journal of Remote Sensing vol 17 no 10 pp 1907ndash1914 1996

[15] L Chaturvedi K R Carver J C Harlan G D Hancock FV Small and K J Dalstead ldquoMultispectral remote sensingof saline seepsrdquo IEEE Transactions on Geoscience and RemoteSensing vol 21 no 3 pp 239ndash251 1982

[16] R P Singh and S K Srivastav ldquoMapping of waterlogged andsalt-affected soils using microwave radiometersrdquo InternationalJournal of Remote Sensing vol 11 no 10 pp 1879ndash1887 1990

[17] M Menenti A Lorkeers and M Vissers ldquoAn applicationof thematic mapper data in Tunisiardquo International Journal ofApplied Earth Observation vol 1 pp 35ndash42 1986

[18] M A Mulders and G F Epema ldquoThe thematic mapper a newtool for soil mapping in arid areasrdquo International Journal ofApplied Earth Observation vol 1 pp 24ndash29 1986

[19] S K Saha M Kudrat and S K Bhan ldquoDigital processing ofLandsat TM data for wasteland mapping in parts of AligarhDistrict (Uttar Pradesh) Indiardquo International Journal of RemoteSensing vol 11 no 3 pp 485ndash492 1990

[20] B R Rao R S Dwivedi L Venkataratnam et al ldquoMapping themagnitude of sodicity in part of Indo-Gangetic plains of UttarPradesh Northern India using Landsat datardquo InternationalJournal of Remote Sensing vol 12 no 3 pp 1419ndash1425 1991

[21] M Y Naseri Characterization of salt-affected soils for modelingsustainable land management in the semi-arid environment acase study in the Gorgan region Northeast Iran [PhD thesis]Ghent University Ghent Belgium 1998


[22] A A Darvishsefat M H Damavandi M Jafari and G RZehtabiyan ldquoAssessing of Landsat TM images for using in soilsalinity classificationrdquo Journal of Desert vol 5 no 2 2000

[23] T Zhang G Zhao C Chang et al ldquoInformation extractionmethod of soil salinity in typical areas of the yellow river deltabased on landsat imageryrdquo Agricultural Sciences vol 06 no 01pp 71ndash77 2015

[24] L M Montandon and E E Small ldquoThe impact of soilreflectance on the quantification of the green vegetation fractionfrom NDVIrdquo Remote Sensing of Environment vol 112 no 4 pp1835ndash1845 2008

[25] B DWardlow and S L Egbert ldquoLarge-area cropmapping usingtime series MODIS 250mNDVI data an assessment for the USCentral Great Plainsrdquo Remote Sensing of Environment vol 112no 3 pp 1096ndash1116 2008

[26] Z P Yang J X Gao C P Zhou et al ldquoSpatio-temporalchanges of NDVI and its relation with climatic variables inthe source regions of the Yangtze and Yellow riversrdquo Journal ofGeographical Sciences vol 21 no 6 pp 979ndash993 2011

[27] A E K Douaoui H Nicolas and CWalter ldquoDetecting salinityhazards within a semiarid context by means of combining soiland remote-sensing datardquo Geoderma vol 134 no 1-2 pp 217ndash230 2006

[28] G Jiapaer X Chen and A M Bao ldquoA comparison of methodsfor estimating fractional vegetation cover in arid regionsrdquoAgricultural and Forest Meteorology vol 151 no 12 pp 1698ndash1710 2011

[29] B R M Rao R S Dwivedi L Venkataratnam et al ldquoMappingthe magnitude of sodicity in part of the Indo-Gangetic plainsof Uttar Pradesh northern India using Landsat-TM datardquoInternational Journal of Remote Sensing vol 12 no 3 pp 419ndash425 1991

[30] A Srivastava N K Tripathi and K V G K Gokhale ldquoMap-ping groundwater salinity using IRS-1B LISS II data and GIStechniquesrdquo International Journal of Remote Sensing vol 18 no13 pp 2853ndash2862 1997

[31] R S Dwivedi and K Sreenivas ldquoDelineation of salt-affectedsoils and waterlogged areas in the Indo-Gangetic plains usingIRS-1C LISS-III datardquo International Journal of Remote Sensingvol 19 no 14 pp 2739ndash2751 1998

[32] M Elhag and J Bahrawi ldquoCloud coverage disruption forgroundwater recharge improvement using remote sensing tech-niques in Asir region Saudi Arabiardquo Life Science Journal vol 11no 1 pp 192ndash200 2014

[33] R Goossens M De Dapper A Gad and Th Ghabour ldquoAmodel for monitoring and prediction of soil salinity andwaterlogging in the Ismailia area (Egypt) based on remotesensing and GISrdquo in Proceedings of the International Symposiumon Operationalization of Remote Sensing vol 6 pp 97ndash107 ITCEnschede The Netherlands April 1993

[34] S Casas ldquoSalinity assessment based on combined use of remotesensing and GISrdquo in Use of Remote Sensing Techniques inIrrigation and Drainage Unknown pp 141ndash150 FAO RomeItaly 1995

[35] R Shaw ldquoEstimation of the electrical conductivity of saturationextracts from the electrical conductivity of 15 soil watersuspensions and various soil propertiesrdquo Project Report SeriesQO94025 Department of Primary Industries QueenslandAustralia 1994

[36] P S Chavez Jr ldquoImage-based atmospheric correctionsmdashrevisited and improvedrdquo Photogrammetric Engineering andRemote Sensing vol 62 no 9 pp 1025ndash1036 1996

[37] U Beisl J Telaar and M Schonermark ldquoAtmospheric correc-tion reflectance calibration and BRDF correction for ADS40image datardquo in Proceedings of the International Archives ofthe Photogrammetry Remote Sensing and Spatial InformationSciences vol 87 part B7 pp 7ndash12 Beijing China 2008

[38] G E P Box ldquoSome theorems on quadratic forms applied in thestudy of analysis of variance problems II Effects of inequalityof variance and of correlation between errors in the two-wayclassificationrdquo Annals of Mathematical Statistics vol 25 no 3pp 484ndash498 1954

[39] A Dehni and M Lounis ldquoRemote sensing techniques forsalt affected soil mapping application to the Oran Region ofAlgeriardquo Procedia Engineering vol 33 pp 188ndash198 2012

[40] A Allbed and L Kumar ldquoSoil salinity mapping and monitoringin arid and semi-arid regions using remote sensing technologya reviewrdquo Advances in Remote Sensing vol 02 no 04 pp 373ndash385 2013

[41] A K Koshal ldquoSpectral characteristics of soil salinity areas inparts of South-West Punjab through remote sensing and GISrdquoInternational Journal of Remote Sensing and GIS vol 1 no 2 pp84ndash89 2012

[42] L Lei T Tiyip J-L Ding H-N Jiang and A KelimuldquoStudy on the soil salinization monitoring based on measuredhyperspectral andHSI datardquo Spectroscopy and Spectral Analysisvol 34 no 7 pp 1948ndash1953 2014

[43] D J Major F Baret and G Guyot ldquoA ratio vegetation indexadjusted for soil brightnessrdquo International Journal of RemoteSensing vol 11 no 5 pp 727ndash740 1990

[44] N M Khan V V Rastoskuev Y Sato and S ShiozawaldquoAssessment of hydrosaline land degradation by using a simpleapproach of remote sensing indicatorsrdquo Agricultural WaterManagement vol 77 no 1ndash3 pp 96ndash109 2005

[45] A Bannari A M Guedon A El-Harti F Z Cherkaoui and AEl-Ghmari ldquoCharacterization of slightly and moderately salineand sodic soils in irrigated agricultural land using simulateddata of advanced land imaging (EO-1) sensorrdquoCommunicationsin Soil Science and Plant Analysis vol 39 no 19-20 pp 2795ndash2811 2008

[46] A Abbas and S Khan ldquoUsing remote sensing techniquesfor appraisal of irrigated soil salinityrdquo in Proceedings of theInternational Congress on Modelling and Simulation (MODSIMrsquo07) L Oxley and D Kulasiri Eds pp 2632ndash2638 Modellingand Simulation Society ofAustralia andNewZealand BrightonUK December 2007

[47] C W Zhang J K Tang X J Yu C L Wang and S J MildquoQuantitative retrieval of soil salt content based on remotesensing in theYellowRiver deltardquo Journal of GraduateUniversityof Chinese Academy of Sciences vol 30 pp 220ndash227 2013

[48] J-L Ding M-C Wu H-X Liu and Z-G Li ldquoStudy on thesoil salinization monitoring based on synthetical hyperspectralindexrdquo Spectroscopy and Spectral Analysis vol 32 no 7 pp1918ndash1922 2012

[49] D Abdellatif and L Mourad ldquoRemote sensing techniques forsalt affected soil mapping application to the Oran Region ofAlgeriardquo Procedia Engineering vol 33 pp 188ndash198 2012

[50] K Krivoruchko A Gribov and E Krause ldquoMultivariate arealinterpolation for continuous and count datardquo Procedia Environ-mental Sciences vol 3 pp 14ndash19 2011

[51] Soil Survey Staff Keys to Soil Taxonomy United States Depart-ment of Agriculture Natural Resources Conservation Service10th edition 2006


[52] B Jiang ldquoHeadtail breaks a new classification scheme for datawith a heavy-tailed distributionrdquo The Professional Geographervol 65 no 3 pp 482ndash494 2013

[53] D R Tilley M Ahmed J H Son and H BadrinarayananldquoHyperspectral reflectance response of freshwater macrophytesto salinity in a brackish subtropical marshrdquo Journal of Environ-mental Quality vol 36 no 3 pp 780ndash789 2007

[54] G A Carter ldquoResponses of leaf spectral reflectance to plantstressrdquo American Journal of Botany vol 80 no 3 pp 239ndash2431993

[55] A Psilovikos and M Elhag ldquoForecasting of remotely senseddaily evapotranspiration data over Nile Delta region EgyptrdquoWater Resources Management vol 27 no 12 pp 4115ndash41302013

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Active and Passive Electronic Components

Control Scienceand Engineering

Journal of



RotatingMachinery


Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics


Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of


Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation



DistributedSensor Networks



composition of TMbands (2 3 and 4) to distinguish betweenmoderate and high soil content soils Naseri [21] used bothunsupervised and supervised classification algorithms todelineate soil salinity on a regional scale using three bandsof MSS and four bands of TM Darvishsefat et al [22] usedprincipal component analysis to map soil salinity using fourbands of ETM+ (3 4 5 and 7) Zhang et al [23] used differentbands of OLI-8 to develop a soil salinity detection algorithmbased on the Soil Salinity Information Extraction (SSIE)

Implementation of remote sensing data in soil salinitymapping is firm to employ since salinization is not a staticprocessTherefore a number of derivatives and alternatives tonormalized difference vegetation index have been anticipatedto discourse soils salinization monitoring and mapping [24ndash26] Different remote sensing indices such as the salinityindex (SI) the normalized difference salinity index (NDSI)and the brightness index (BI) in addition to normalizeddifference vegetation index (NDVI) were implemented toinvestigate the means by which these indices work for soilsalinity mapping in the arid environment [27 28] Thesalinity index (SI) which combines the blue and red bands issensitive to the surface reflectance of salt-affected land withsparse vegetation cover [27]

The perspective of using remote sensing data and GISpractices has been demonstrated in several scholarly worksto be the most resourceful [29ndash32] The comprehensive useof satellite remote sensing and GIS has been recognized tobe a cost-effective method for monitoring soils salinizationin poorly drained basins [33 34] However state-of-the-artoutcome specifies that there is no ideal combination of datatype and questioning method which can be functional withidentical realization across approximately several environ-mental conditions [25ndash27]

The aim of current research project is to differentiatethe salinized soils from nonsalinized soils implementingdifferent approaches of digital image classification and bandmath techniques To reach that objective several soil salinityindices were applied to the Landsat 8 data verified withground truth data to validate the percentage of estimationsaccuracy based on different errorrsquos assessment approach

2 Materials and Methods

21 Study Area The study area Wadi Al Dawasir townis located in the plateau of Najd at latitude of 44∘431015840 andlongitude of 20∘291015840 about 300 km south of the capital cityRiyadh (Figure 1) This study area is comprised of gravellytableland disconnected by insignificant sandy oases andisolated mountain bundles Across the Arabian Peninsulaas a whole the tableland slopes toward the east from anelevation of 1360 meters in the west to 750 meters at itseasternmost limit Wadi Al Dawasir and Najran regions arethemajor irrigationwater abstraction fromAl-Wajid AquiferAgriculture in Wadi Al Dawasir area consists of technicallyhighly developed farm enterprises that operate modern pivotirrigation system The size of center pivot ranges from 30 hato 60 ha with farms managing hundreds of pivots withthe corresponding number of wells The main crop grown

Figure 1 Location of the study area

in winter is wheat and occasionally potatoes tomatoes ormelons All year fodder consists of alfalfa which is cut up to10 times a year for food Typical summer crops for fodder aresorghum and Rhodes grass which is perennial but dormantin winter The shallow alluvial aquifers could not sustainthe high groundwater abstraction rates for a long time andgroundwater level declined dramatically in most areas Mete-orological features of the area are speckled Five elements ofmeteorology are constantly recorded through fixed weatherstation located within the study area Temperature variesfrom 6∘C as minimum temperature to 43∘C as maximumtemperature Relative humidity is mostly stable at 24 Solarradiation of average sunrise duration is generally 11 hrsdayAverage wind speed is closer to 13 kmhr and may reach upto 46 kmhr in thunderstorm incidents Finally mean annualrainfall is about 376mm [5]

22Methodological Framework The current research work isbased on assessing a regression correlation between groundtruth data and its corresponding salinity indexrsquos valuesconducted from satellite images Therefore accurate syn-chronization of ground truth data collection and satellitebypassing were exercised

221 Soil Sampling and Analysis A total number of 150random soil samples were collected from Wadi Al Dawasirarea with a minimum distance of 100m between the sampleslocations to avoid data clumping (Figure 2) The standardprocedure of determining soil extract salinity in terms ofElectrical Conductivity (EC) was followed according to Shaw[35] under laboratory condition

222 Remote Sensing Data Enhancement Remote sensingdata was obtained from Landsat Operational Land Imager(OLI-8) which was acquired on April 17th 2014 Typi-cal atmospheric and radiometric corrections and spatialresolution enhancement were implemented for each bandindividually Furthermore overcorrection was analyzed forthe red and near infrared bands [36] Atmospheric correctionis basically considering blue band offset Therefore thelargest atmospheric consequence is theoretical to be the mosttruthful value within the blue band The calibrated radiancevalue of this offset allows deciding 120581 valueTherefore 120582minus120581 rule


0 125 25 50

(km)

N

WS

E







y = 25315x + 18571

10

20

30

40

0 2 4 6 8

Estim

ated

soil

salin

ity v

alue

s (pp

m)

Thou

sand

s


R2 = 09764












[43]






[45]


[45]


[45]


[46]


[46]





10

05

00

minus05

minus10

Component 1 (976)













10

3552

3

LowModerate

HighExtreme


0 25 50

(km)Soil salinity

LowModerate

HighExtreme

N

W

S

E


0

10

20

30

40

50


Tota

l agr

icul

tura

l are

a (

)

Soil salinity class




4 Conclusions






Acknowledgment


References




























































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










0 125 25 50

(km)

N

WS

E







y = 25315x + 18571

10

20

30

40

0 2 4 6 8

Estim

ated

soil

salin

ity v

alue

s (pp

m)

Thou

sand

s


R2 = 09764












[43]






[45]


[45]


[45]


[46]


[46]





10

05

00

minus05

minus10

Component 1 (976)













10

3552

3

LowModerate

HighExtreme


0 25 50

(km)Soil salinity

LowModerate

HighExtreme

N

W

S

E


0

10

20

30

40

50


Tota

l agr

icul

tura

l are

a (

)

Soil salinity class




4 Conclusions






Acknowledgment


References




























































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation













[43]






[45]


[45]


[45]


[46]


[46]





10

05

00

minus05

minus10

Component 1 (976)













10

3552

3

LowModerate

HighExtreme


0 25 50

(km)Soil salinity

LowModerate

HighExtreme

N

W

S

E


0

10

20

30

40

50


Tota

l agr

icul

tura

l are

a (

)

Soil salinity class




4 Conclusions






Acknowledgment


References




























































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation














10

3552

3

LowModerate

HighExtreme


0 25 50

(km)Soil salinity

LowModerate

HighExtreme

N

W

S

E


0

10

20

30

40

50


Tota

l agr

icul

tura

l are

a (

)

Soil salinity class




4 Conclusions






Acknowledgment


References




























































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation













Acknowledgment


References




























































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation















































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation
















RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation











RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation









Documents

Research Article Evaluation of Different Soil Salinity ...downloads.hindawi.com/journals/js/2016/7596175.pdfResearch Article Evaluation of Different Soil Salinity Mapping Using Remote