Lake Area Analysis Using Exponential Smoothing Model and Long Time-Series Landsat Images in Wuhan, China

sustainability

Article Lake Area Analysis Using Exponential Smoothing Model and Long Time-Series Landsat Images in Wuhan, China

Gonghao Duan 1,* ID and Ruiqing Niu 2

1 Faculty of Information Engineering, China University of Geosciences, Wuhan 430074, China 2 College of Geophysics and Geomatics, China University of Geosciences, Wuhan 430074, China; [email protected] * Correspondence: [email protected]

Received: 14 November 2017; Accepted: 8 January 2018; Published: 9 January 2018

Abstract: The loss of lake area significantly influences the climate change in a region, and this loss represents a serious and unavoidable challenge to maintaining ecological sustainability under the circumstances of lakes that are being filled. Therefore, mapping and forecasting changes in the lake is critical for protecting the environment and mitigating ecological problems in the urban district. We created an accessible map displaying area changes for 82 lakes in the Wuhan city using remote sensing data in conjunction with visual interpretation by combining field data with Landsat 2/5/7/8 Thematic Mapper (TM) time-series images for the period 1987–2013. In addition, we applied a quadratic exponential smoothing model to forecast lake area changes in Wuhan city. The map provides, for the first time, estimates of lake development in Wuhan using data required for local-scale studies. The model predicted a lake area reduction of 18.494 km2 in 2015. The average error reached 0.23 with a correlation coefficient of 0.98, indicating that the model is reliable. The paper provided a numerical analysis and forecasting method to provide a better understanding of lake area changes. The modeling and mapping results can help assess aquatic habitat suitability and property planning for Wuhan lakes.

Keywords: lake; forecasting; mapping; quadratic exponential

1. Introduction With the development of remote sensing technology, increasingly large data sets of imagery have been used for water resource assessment [1,2]. The Hubei Province is known as “the province of a thousand lakes” in China, as there are more than 80 shallow lakes densely distributed throughout the Wuhan development area alone. The loss of lakes significantly influences the level of environmental pollution as well as climate change in a region, and this loss represents a serious and unavoidable challenge to maintaining ecological sustainability under the circumstances of lakes that are being filled. Therefore, mapping changes in lake areas is critical for protecting the environment and mitigating ecological problems in the Wuhan urban area. Several mapping water studies have been conducted for different research fields. They noted the importance of wetlands for urban landscape planning and delineated the relationship between human health and urban wetland environmental ecosystems [3,4]. Researchers quantified the changes in both lake areas and boundaries from multiple sets of remote sensing images [5,6]. Satellite remote sensing plays a significant role in water resource surveying, hydrological monitoring, wetland protection, and geological disaster warning systems because of the periodic repetition of data and the ability to cover a large area [7–10]. Experts obtained the modified normalized difference water index (MNDWI) by analyzing the differences in the amounts of radiation received from water bodies

Sustainability 2018, 10, 149; doi:10.3390/su10010149 www.mdpi.com/journal/sustainability SustainabilitySustainability 20182018,, 1010,, 149149 2 of 16

normalized difference water index (MNDWI) by analyzing the differences in the amounts of andradiation non-water received features from [ 11water]. Some bodies experts and madenon-water a comparative features [11]. study Some of automated experts made water a classiﬁcationcomparative methodsstudy of automated [12] and used water moderate-resolution classification methods sensor [12] imagery, and used such moderate-resolution as that acquired from sensor the imagery, Landsat Thematicsuch as that Mapper acquired (TM) from and the Moderate Landsat Themat Resolutionic Mapper Imaging (TM) Spectroradiometer and Moderate Resolution (MODIS) Imaging sensors, toSpectroradiometer obtain the distributions (MODIS) of desiredsensors, features to obtain among the distributions various study of areas desired [13– 15features]. Many among investigations various primarilystudy areas utilized [13–15]. remote Many sensing investigations images primarily from the lastutilized 10 years remote to exploresensing changesimages from in lake the areas last or10 otheryears featuresto explore to studychanges developmental in lake areas changingor other features trends resulting to study from developmental human interaction changing with trends the environmentresulting from[16 human–22]. The interaction lakes in thewith Wuhan the environment urban district [16–22]. are taken The as lakes research in the subjects Wuhan herein urban to applydistrict a are novel taken approach as research [23– 25subjects] to map herein lake to changes apply a overnovel a approach nearly 30-year [23–25] period. to map We lake utilize changes the MNDWIover a nearly and the30-year relationships period. We among utilize various the MN remoteDWI sensing and the bands relationships to extract among the spatial various information remote ofsensing lake surfaces bands betweento extract 1987 the and spatial 2013. Furthermore, information we of provide lake surfaces a valid exponential between model1987 and to forecast 2013. futureFurthermore, changes. we The provide set of a measures valid exponential used herein model includes to forecast the degree future of changes. model ﬁt, The the set robustness of measures of theused model, herein and includes the prediction the degree skill of [26model–29]. fit, Finally, the robustness we map the of changes the model, in lake and area the throughoutprediction skill the Wuhan[26–29]. urban Finally, development we map the district changes (1987–2013), in lake area the throughout results of the which Wuhan are readily urban explainabledevelopment and district may possess(1987–2013), practical the results applications. of which are readily explainable and may possess practical applications.

2. Study Area The city of Wuhan isis locatedlocated toto thethe easteast ofof JianghanJianghan PlainPlain andand isis situatedsituated inin thethe middle of China 2 withwith aa totaltotal areaarea ofof 8494.418494.41 kmkm2.. The The city city is divided into three parts by twotwo majormajor riversrivers thatthat cutcut throughthrough thethe city city centre: centre: the the Yangtze Yangtze River, River, which which is the is third the largestthird largest river in river the world in the and world the Hanjiang and the RiverHanjiang (the River Yangtze (the River’s Yangtze largest River’s tributary). largest tributar The collectivey). The collective water surface water of surface the two of rivers, the two covering rivers, 2 upcovering to 2217.6 up kmto 2217.6, accounts km2, for accounts nearly afor quarter nearly of a thequarter urban of distinct, the urban which distinct, is the which highest is among the highest all of China’samong all major of China’s cities (Figure major 1citiesa). (Figure 1a).

(a) (b)

Figure 1. 1.( a()a) represents represents the the newest newest Wuhan Wuhan urban urban boundary boundary shapeshape ﬁlefile thatthat wewe employemploy toto covercover thethe Landsat 88 remoteremote sensingsensing image image (2013); (2013); (b ()b shows) shows a relativelya relatively scattered scattered distribution distribution of lakes,of lakes, which which are mainlyare mainly concentrated concentrated within within central central and and southern southern Wuhan. Wuhan. The The lakes lakes are are more morelikely likely toto appearappear in low-lyinglow-lying areas with higher elevations, and many small lakes are are developed around around the the main main river. river. TheThe twotwo largestlargest urbanurban lakes lakes (East (East Lake Lake and and Tangxun Tangxun Lake) Lake) are are located located in thein the middle-west middle-west section section of the of studythe study area. area. This This lake distributionlake distribution shows shows that there that arethere abundant are abundant water resourceswater resources in the urban in the area. urban area. As of 2015, there were approximately 166 lakes in Wuhan, which is known as the “city of the As of 2015, there were approximately 166 lakes in Wuhan, which is known as the “city of the hundred lakes”. The lake surface area was 803.17 km2 at normal water level. Tangxun Lake (47.6 km2) hundred lakes”. The lake surface area was 803.17 km2 at normal water level. Tangxun Lake (47.6 is the largest city-lake in China at present [30]. Many large lakes are situated on either side of the two km2) is the largest city-lake in China at present [30]. Many large lakes are situated on either side of main rivers, forming a network of lakes (Figure1b). Our research team conducted a thorough study of the two main rivers, forming a network of lakes (Figure 1b). Our research team conducted a the 82 main lakes in the Wuhan urban area by means of follow-up methods. thorough study of the 82 main lakes in the Wuhan urban area by means of follow-up methods. Sustainability 2018, 10, 149 3 of 16

Sustainability 2018, 10, 149 3 of 16 In our research, we focus on annual changes in the body of the main lakes (1987–2013). In our research, we focus on annual changes in the body of the main lakes (1987–2013). The Theinformation information of ofthe the lake lake boundaries boundaries is isextrac extractedted from from remote remote sensing sensing data, data, including including the the characteristicscharacteristics of of the the electromagnetic electromagnetic radiationradiation from earth earth objects. objects. As As a aresult, result, we we provide provide a priori a priori knowledgeknowledge for for the the protection protection of of water water resource resource environments environments in in Wuhan. Wuhan. In In addition addition to to techniques techniques for thefor extraction the extraction of lake of areaslake areas using, using, for example, for exampl thee, MNDWI, the MNDWI, we alsowe also conducted conducted ﬁeld field trips trips to every to lake.every Figure lake.2 showsFigure the2 shows ﬁeld conditionsthe field conditions of the South of Princethe South Lake. Prince More Lake. details More will details be discussed will be in thediscussed Results andin the Discussion Results and sections. Discussion sections.

Figure 2. Field photographs depicting South Prince Lake in 2015 (a); Residential buildings and Figure 2. Field photographs depicting South Prince Lake in 2015 (a); Residential buildings and communities are observed near the lake. The area was well retrieved in Landsat 8 imagery in 2013 (b) communities are observed near the lake. The area was well retrieved in Landsat 8 imagery in 2013 (b) and in Landsat 5 imagery in 2001 (c); The water area in 2013 (b) was slightly decreased due to real and in Landsat 5 imagery in 2001 (c); The water area in 2013 (b) was slightly decreased due to real estate projects in the southeastern corner, and some parts were filled for industrial use. estate projects in the southeastern corner, and some parts were ﬁlled for industrial use. 3. Materials and Methods 3. Materials and Methods Lakes area change were a comprehensive result of many external factors, so our research did notLakes focus areaon any change external were factors a comprehensive affecting lake resultincrease of or many loss, external such as factors,climate changes, so our research and only did notaddresses focus on the any general external lake factors changes affecting in Wuhan. lake We increase organized or loss, the such images as climatetaken during changes, the andperiod only addressesfrom 1987 the to general 2013 in lakeconsideration changes inof Wuhan.the acquisition We organized time, coverage, the images cloudage, taken and during whether the period resolved from 1987stripes to 2013 were in available. consideration Then, of we the overlay acquisition the images time,coverage, over the Wuhan cloudage, urban and area whether so that resolved the selected stripes weredata available. can fully cover Then, the we research overlay area. the imagesWe choose over im theages Wuhan such that urban their areatotal socloudage that the was selected less than data can10%. fully This cover step the is researchimportant area. because We choose excessive images cloud such cover that can their obscure total cloudagewater bodies was lessand thaninduce 10%. Thisdifficulties step is important in subsequent because extraction excessive tasks. cloud cover can obscure water bodies and induce difﬁculties in subsequentBased on extraction the above tasks. considerations, we choose over 10 periods of images as a data set spanning from 1987 to 2013 to study variations in the major lakes of the Wuhan urban area. The remote Sustainability 2018, 10, 149 4 of 16

Based on the above considerations, we choose over 10 periods of images as a data set spanning from 1987 to 2013 to study variations in the major lakes of the Wuhan urban area. The remote sensing images used in this paper were all acquired from the Landsat satellites, and the data were obtained from either a cloud-based database of geographic information from the Chinese Academy of Sciences or theSustainability United States2018, 10, Geological149 Survey (USGS) website. Our research takes advantage of the4 of 16 above data to study the linear and polygonal layers of lakes primarily through a water extraction method in additionsensing images to visual used interpretation in this paper withinwere all ArcGIS.acquired Thefrom main the Landsat image satellites, data are and presented the data in were Table 1, and aobtained sample from well-processed either a cloud-based image is database shown inof Figuregeographic3. information from the Chinese Academy of Sciences or the United States Geological Survey (USGS) website. Our research takes advantage of the above data to study the linear and polygonalTable 1. Image layers data of lakes set. primarily through a water extraction method in addition to visual interpretation within ArcGIS. The main image data are presented in SatelliteTable 1, and Sensor a sample Resolutionwell-processed (m) image is Data shown Identiﬁcation in Figure 3. Date Cloudage (%)

Landsat 4 TM 30 mTable LT41230391989042XXX02 1. Image data set. 1989/02/11 None LT51230391987253BJC00 1987/9/10 4.7% Satellite Sensor Resolution (m) Data Identification Date Cloudage (%) Landsat 4 TM 30 m LT51230391994272BJC00 LT41230391989042XXX02 1989/02/11 1994/9/29 None None Landsat 5 TM 30 m LT51230392001259BJC00LT51230391987253BJC00 1987/9/10 2001/9/16 4.7% 6.7% LT51230392005254BJC00LT51230391994272BJC00 1994/9/29 2005/9/11 None 9% Landsat 5 TM 30 m LT51230392009249BJC00LT51230392001259BJC00 2001/9/16 2009/9/6 6.7% None LT51230392005254BJC00 2005/9/11 9% LE71230392003273EDC01 2003/9/30 6% LT51230392009249BJC00 2009/9/6 None Landsat 7 ETM+ 30/15 m LE71230392007332EDC00LE71230392003273EDC01 2003/9/30 2007/11/28 6% None Landsat 7 ETM+ 30/15 m LE71230392011343EDC00LE71230392007332EDC00 2007/11/28 2011/12/9 None None Landsat 8 OLI 30/15 m LC81230392013164LE71230392011343EDC00 LGN00 2011/12/9 2013/6/13 None None Landsat 8 OLI 30/15 m LC81230392013164 LGN00 2013/6/13 None

Figure 3. Operational Land Imager (OLI) remote sensing image (Wuhan urban area, Landsat 8, 13 Figure 3. Operational Land Imager (OLI) remote sensing image (Wuhan urban area, Landsat 8, June 2013) that was processed using radiation correction, atmospheric calibration, area shape cutting 13 Juneand 2013 strip) repair that was work. processed using radiation correction, atmospheric calibration, area shape cutting and strip repair work. 3.1. Water Extraction 3.1. Water Extraction The images used herein consist of two types of data organisation schemes. The range of grey valuesThe images of the TM used and herein Enhanced consist TM of Plus two (ETM+) types ofsensors data is organisation between 0 and schemes. 255 (pre-2013) The range and the of grey valuesgrey of values the TM of and the EnhancedLandsat 8 TMOperational Plus (ETM+) Land sensorsImager is(OLI) between sensor 0 andrange 255 from (pre-2013) 0 to 65,535 and the grey(post-2013). values of the With Landsat the 8same Operational image resolution, Land Imager the (OLI)information sensor rangein the from Landsat-8 0 to 65,535 imagery (post-2013). is undoubtedly greater [31]. Therefore, we adopted two different methods for the water extraction under two separate conditions. Image classification and feature extraction processes are complex and may be affected by many factors [32]. The brightness value of water is different from those of shoals and shaded relief but is Sustainability 2018, 10, 149 5 of 16

With the same image resolution, the information in the Landsat-8 imagery is undoubtedly greater [31]. Therefore, we adopted two different methods for the water extraction under two separate conditions. Image classification and feature extraction processes are complex and may be affected by many factors [32]. The brightness value of water is different from those of shoals and shaded relief but is similar to those of settlements or woodlands in bands 2 and 3 of TM and ETM+ images, and it is less than those of other surface features in bands 4 and 5. With these reflection characteristics, we can effectively distinguish between water and other surface features by choosing a specific band combination. In addition, we can exclude the effects of interference from shaded relief and snow by setting a threshold. In this paper, we select bands 3, 4, 5 and 6 to perform the band operations because there is a broad range of image brightness values in OLI data. Moreover, the ratios of the brightness values in bands 3 and 4 and in bands 5 and 6, which effectively represent water bodies, will be amplified. The model we use in Landsat 8 can be written as follows:

(OLI3 + OLI4)/(OLI5 + OLI6) > 1, (1)

Normally, water bodies are characterized by (TM2 + TM3) > (TM4 + TM5) [33]. We calculate the brightness of an image pixel by placing a strong reﬂection band in the numerator and a weak reﬂection band in the denominator, following which their contrast will be further expanded. Accordingly, the brightness of the studied feature in the resulting image will be the maximum and the other features will be correspondingly suppressed, which will make the water surface easily differentiable from background objects. The MNDWI [11] that we employ for the TM and ETM+ images (pre-2013) can be written as follows: (p(Green) − p(MIR))/(p(Green) + p(MIR)) > 0.44, (2)

3.2. The Quadratic Exponential Smoothing Model The extraction of the water surface may reﬂect changes in the lake area at a given time, but future changes in the lake area are often required for more thoughtful measures. An exponential smoothing model, which is a special type of weighted average model, is sensitive to historical data that is close to the forecasting period, and the weight of the model declines with increasing time [34,35]. The exponential smoothing model requires fewer training data and less modeling time, lake area generally has a decreasing tendency and historical relevance year by year, so this model is therefore more suitable for evaluating complex variations in lake surface areas than other models. In addition, when the trend of the data is either rising or falling, the quadratic exponential smoothing method will be more effective because the single exponential method may output large forecast deviations [36]. The quadratic exponential smoothing method applies exponential smoothing based on the single exponential smoothing model by simultaneously considering the overall trend of the curve [37]. In most cases, we only require three sets of input training data and a parameter value that can be utilized to establish the forecast model. As lake areas are essentially decreasing on an annual basis, this algorithm is well adapted for predicting lake areas. The linear quadratic exponential smoothing method is described using the following equation:

(1) (1) St = aYt−1 + (1 − a)St−1 , (3)

(2) (1) (2) St = aSt + (1 − a)St−1 , (4)

In this paper, St and St−1 represent the smoothing values for the lake areas at times t and t − 1, respectively, and a is the smoothing coefﬁcient, namely, the damping coefﬁcient [38]. Yt−1 is the actual (1) (2) lake area at t − 1. The values of St and St are known, and the quadratic exponential smoothing model can be written as follows: Ft+T = dt + btT, (5) Sustainability 2018, 10, 149 6 of 16

(1) (2) dt = 2 St − St , (6) (1) (2) bt = a(St − St )/(1 − a), (7)

T denotes the predictive periods, Ft+T represents the predicted value of the lake area at the time t + T, and both dt and bt are parameters of the data at the time T. It is not difficult to discern Sustainability 2018, 10, 149 6 of 16 from Formulaes (3)–(7) that the value of the smoothing parameter a plays a crucial role, as it reflects theapproximation approximation of the of thepredicted predicted value value with with respec respectt to to the the current current data data value value [[39].39]. In practicalpractical applications, the the smoothing smoothing parameter parameter is is a aconstant constant between between 0 0and and 1. 1.When When a approachesa approaches 1, the 1, thepredicted predicted values values will will be becloser closer to tothe the current current value value and and far far from from the the smoothness smoothness value value of thethe historical data, data, and and the the slope slope of ofthe the preceding preceding curve curve will will be changing be changing faster faster prior priorto becoming to becoming steep. steep.When a When approachesa approaches 0, the predicted 0, the predicted value will value be cl willoser be to closerthe smoothness to the smoothness value of the value historical of the historicaldata, which data, means which the means curve the will curve be willsmoother be smoother than before than before(Figure (Figure 4). In4 ).previous In previous studies, studies, the thesmoothing smoothing coefficient coefficient was was generally generally based based on on the the pr principleinciple of of the the minimum minimum me meanan square square deviation, deviation, and experiencedexperienced methodsmethods were were often often utilised utilised to to confirm confirm the the values values of theof smoothnessthe smoothness parameters parameters [40]. Our[40]. researchOur research considers considers the characteristics the characteristics of lake of areaslake areas affected affected by annual by annual variations, variations, and we and have we determinedhave determined that 0.9 that is the0.9 is optimal the optimal smoothness smoothness coefficient coefficient after conductingafter conducting many many tests. tests.

Figure 4. Variation of St with smoothing parameter. We use an incremental sequence as an example, Figure 4. Variation of St with smoothing parameter. We use an incremental sequence as an example, and the predicted values (St) will be closer to the current value and far from the smoothness value of and the predicted values (St) will be closer to the current value and far from the smoothness value of thethe historicalhistorical data,data, thethe slopeslope ofof thethe precedingpreceding curvecurve willwill bebe changingchanging fasterfaster priorprior toto becomingbecoming steep.steep.

In addition, some experts used the mean absolute error (MAE) and correlation coefficient (R) to MAE R representIn addition, the prediction some experts accuracy used of the the mean quadrati absolutec exponential error ( smoothing) and correlation model [41], coefﬁcient the formulae ( ) to representfor which theare predictiongiven as follows: accuracy of the quadratic exponential smoothing model [41], the formulae for which are given as follows: ∧ 1= 1 −∧ , MAEMAE= Y(Yt)(t−) YY((t) , (8)(8) n ∑n

−− −− [([(AA−i A)(A)(B B−i B)]B )] R == ∑ i i R r r , (9)(9) (A −−A)2 × (B − B )2  i 2  i 2 ∑ (Ai − A) × ∑ (Bi − B) , Y(t) is the actual value while the other variable represents the predicted value of the model, and n is the total number of data in Formula (8). In Formula (9), Ai and Bi represent the actual value and the predicted value of the model, the other variable represents the mean of the two values. Smaller MAE values indicate smaller model errors, while larger R values represent higher correlations between the fitted data and actual values, thereby suggesting that the model has a better forecasting ability. Considering the effect of overfitting in forecasting models for long period cases, our research only predicted lake area in 2015 for a better accuracy and understanding.

Sustainability 2018, 10, 149 7 of 16

Y(t) is the actual value while the other variable represents the predicted value of the model, and n is the total number of data in Formula (8). In Formula (9), Ai and Bi represent the actual value and the predicted value of the model, the other variable represents the mean of the two values. Smaller MAE values indicate smaller model errors, while larger R values represent higher correlations between the ﬁtted data and actual values, thereby suggesting that the model has a better forecasting ability. Considering the effect of overﬁtting in forecasting models for long period cases, our research only predicted lake area in 2015 for a better accuracy and understanding.

3.3. Lake Area Change Index The lake evolutionary history includes body changes, spatial transformations, and quality changes. As areal deformations are first reflected within the imagery, we can initially understand the overall developmental trends of the lakes. The changes in the lake areas and their rates of change may directly reflect the evolution of the ecological environment. Therefore, we took the following equation to describe the extent of the variability in the water bodies:

L1 = Ub − Ua, (10)

L2 = (Ub − Ua)/T, (11)

L1 represents the total range of the variability, L2 denotes the range of annual variations of the lakes, Ua and Ub represent the lake areas at the beginning and end of a study period, and T is the number of years between the starting and ending times. As a result, the area of a lake will increase when L1 > 0 and decrease when L2 < 0. However, for equivalent amplitudes of the total variation or annual variation amplitudes, the actual situations reﬂected within many periods will be different. To obtain a better description of the relative variations among different periods, we use a lake change intensity index to analyse the relative intensity of area changes. The lake change intensity index can be written as follows:

C = 100 × ∆Aba/(Aa × ∆t), (12)

C is the area intensity index of the lake evolution, ∆Aba represents the areal changes within the period from a to b, and Aa is the total area of the lakes during the year a.

4. Results and Discussion

4.1. Lake Area Change Analysis and Forecasting First, we transform the projection according to the Beijing 54 coordinate system for the original remote sensing imagery of the study area and then match the image with a standard geologic map. Subsequently, we use the Wuhan urban development area as a mask to cut out the Landsat 8 remote sensing image. Second, we utilize the MNDWI and the relationships among the sensor bands to distinguish the water bodies from other surface features. We obtained the lake shape data for the study region following binarization and vectorization. Then, after visual interpretation, the spatial location and size of each lake were conﬁrmed, after which we added the generated vector layer of the lake areas as another attribute. Finally, the lakes layer was converted into a raster layer to detect changes. We use the spatial and statistical analysis functions of GIS to calculate the areas of the lakes. The results are shown in the charts of Figure5. To better study the annual variations in the lake areas and to analyse the source of the variations over the two-year periods, we use Formulas (10) and (11) to acquire the total range of variability and the range of annual variations, respectively, for the changes among all of the lakes. The results are shown in Figure6. Sustainability 2018, 10, 149 8 of 16 Sustainability 2018, 10, 149 8 of 16 Sustainabilityecology, 2018and, 10the, 149 reduction in the lake areas became less gradual. According to two main indictors8 of 16 ecology,shown in and Figure the 6,reduction we then calculatedin the lake the areas lake beca changeme less intensity gradual. index According for Wuhan to two (Figure main 7). indictors shown in Figure 6, we then calculated the lake change intensity index for Wuhan (Figure 7). Lake Area (km2) Lake Area (km2) 390.00 370.97 366.17 390.00 361.49 370.00 370.97 356.63 366.17 351.67 361.49 347.01 370.00 356.63 342.15 350.00 351.67 336.50 347.01 350.00 342.15 330.00 336.50319.55 330.00 319.55 310.00 302.89 289.46 310.00 302.89 290.00 278.76 289.46 270.83 290.00 278.76 264.73 270.00 270.83 264.73 270.00 250.00 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 250.00 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 Figure 5. Total area of all of the lakes in the Wuhan urban area (1987–2013). The lake area statistics Figure 5. Total area of all of the lakes in the Wuhan urban area (1987–2013). The lake area statistics Figurefrom every 5. Total two area years of are all compared,of the lakes revealing in the Wuha an ovn erallurban annually area (1987–2013). decreasing The trend. lake With area astatistics decline from every two years are compared, revealing an overall annually decreasing trend. With a decline in fromin the every total areatwo fromyears 370.97 are compared, km2 in 1987 revealing to 264.73 an km ov2erall in 2013, annually a total decreasing decrease of trend. 106.24 With km2 aoccurred decline the total area from 370.97 km2 in 1987 to 264.73 km2 in 2013, a total decrease of 106.24 km2 occurred overin the a total period area of fromtwenty-six 370.97 years.km2 in 1987 to 264.73 km2 in 2013, a total decrease of 106.24 km2 occurred over a period of twenty-six years. over a period of twenty-six years. Total Range of Variability and Range of Annual Variations (km2) 2 0.00 Total Range of Variability and Range of Annual Variations (km ) 0.00 0.00 0.00 -2.00 -1.00 -2.00 -1.00 -4.00 -2.00 -4.00 -2.00 -6.00 -2.40 -2.34 -2.43 -2.48 -2.33 -2.43 -3.00 -2.83 -6.00 -2.40 -2.34 -2.43 -2.48 -2.33 -2.43 -3.05 -3.00 -8.00 -2.83 -4.00 -3.05 -8.00 -3.97 -4.00 -10.00 -5.00 -3.97 -10.00 -5.00 -12.00 -5.35 -6.00 -12.00 -5.35 -6.00 -14.00 -7.00 -6.72 -14.00 -7.00 -16.00 -6.72 -8.00 -16.00 -8.00 -18.00 -8.48 -8.33 -9.00 -18.00 -8.48 -8.33 -9.00 -20.00 -10.00 -20.00 -10.00 Total yearly change Yearly change rangeability Total yearly change Yearly change rangeability Figure 6. Total variations in the lake areas in the Wuhan urban area (1987–2013). Figure 6. Total variations in the lake areas in the Wuhan urban area (1987–2013). The changeFigure intensity 6. Total variationsindex of the in the lakes lake can areas reve in theal a Wuhan clearer, urban more area realistic (1987–2013). meaning than the actualThe change change values intensity for the index lake areasof the in lakes a time can interval. reveal Ina clearer,Figure 7, more the lake realistic change meaning intensity than shows the actualthatThe the change data urban for values lake theperiod change for the between lakeintensity areas 1987 index in a and time maintained 2001 interval. show Invalues that Figure the less 7, average thethan lake −0.827 annual change from rate intensity 1987 of decrease to shows 2001, of 2 thethatwhich, lakes the incompared urban the urban lake with change development the whole intensity observational areas index was maintained approximately period, isvalues a relatively 2.33–2.83 less than weak km−0.827 magnitudeper from year, 1987 of which variation. to 2001, means thatwhich,Between lake conservationcompared 2001 and with 2007, efforts the the whole were intensity effectiveobservational index during increased period, that time. issharply a relatively However, to values weak from rangingmagnitude 2001 to from 2005, of variation.− the2.607 annual to rateBetween−2.219, of decrease which 2001 are rapidlyand 3–4 2007, times increased the thos intensitye tobetween 8.33 index and 1987 increased 8.48 and km 2001.2 persharply The year, intensity to wherein values index ranging more peaked lakes from at needed −2.607 tointo be restored.−2.219, Furthermore,which are 3–4 between times thos 2005e between and 2013, 1987 the and entire 2001. average The intensity lake area index began peaked to decrease at −2.607 annually. in As a result, the variations in the total lake area went through three stages: during 1987–2001, the water area shrank slowly and in a stable manner, and the lakes of Wuhan city were healthy and in good Sustainability 2018, 10, 149 9 of 16 condition; during 2001–2005, the lake surface area diminished at a relatively rapid rate because increasingSustainability numbers 2018, 10, 149 of residential buildings and community projects were constructed near the9 of lakes, 16 following which the ecological environments of the lakes further deteriorated; during 2005–2013, 2003–2005. The variability of the lake areas was dramatic during the next 7 years. Between 2007 and people began to realize the importance of the lakes to the urban ecology, and the reduction in the lake 2013, the intensity index dropped to values between −1.848 and −1.126 in 2011–2013, proving that the areas became less gradual. According to two main indictors shown in Figure6, we then calculated the variation amplitude was clear and that the changes in the lake areas were going to slow down year lake change intensity index for Wuhan (Figure7). by year.

Intensity Index of Lake Evolution (%)

-1.126 2011-2013 -1.424 2009-2011 -1.848 2007-2009 -2.219 2005-2007 -2.607 2003-2005 -2.520 2001-2003 -0.827 1999-2001 -0.700 1997-1999 -0.663 1995-1997 -0.695 1993-1995 -0.672 1991-1993 -0.639 1989-1991 -0.647 1987-1989 -3.000 -2.500 -2.000 -1.500 -1.000 -0.500 0.000

Figure 7. Total changes in the lake areas in the Wuhan urban area (1987–2013). Figure 7. Total changes in the lake areas in the Wuhan urban area (1987–2013). We studied the annual variation in the area of South Prince Lake as an example for the period betweenThe change 1987 and intensity 2013. South index Prince of the lakesLake is can located reveal in a the clearer, middle-west more realistic of the meaningWuhan area than (Figure the actual 8) changeand is values adjacent for theto the lake Yangtze areas in River. a time According interval. Into Figurethe water7, the surface lake changeextracted intensity from the shows remote that thesensing urban lakeimagery, change we can intensity see the index evolution maintained process valuesas presented less than in Figures−0.827 9 fromand 10. 1987 to 2001, which, compared with the whole observational period, is a relatively weak magnitude of variation. Between 2001 and 2007, the intensity index increased sharply to values ranging from −2.607 to −2.219, which are 3–4 times those between 1987 and 2001. The intensity index peaked at −2.607 in 2003–2005. The variability of the lake areas was dramatic during the next 7 years. Between 2007 and 2013, the intensity index dropped to values between −1.848 and −1.126 in 2011–2013, proving that the variation amplitude was clear and that the changes in the lake areas were going to slow down year by year. We studied the annual variation in the area of South Prince Lake as an example for the period between 1987 and 2013. South Prince Lake is located in the middle-west of the Wuhan area (Figure8) and is adjacent to the Yangtze River. According to the water surface extracted from the remote sensing imagery, we can see the evolution process as presented in Figures9 and 10. As shown in Figure 10, the data for each time period were negative among the three indicators, indicating that the overall ecological environment of South Prince Lake was in a state of reduction. The amplitude of variation and the yearly rangeability show that the lake was in an accelerated stage of atrophy between 1987 and 2005 and in a deceleration phase between 2005 and 2013. During the two periods of 1987–1994 and 2011–2013, the yearly rangeability reached −0.46 km2 and −0.31 km2, which does not constitute a large difference. This was perhaps due to slow urban development in 1987–1994, during which time human engineering activity was not substantial. In addition, additional laws and regulations for the preservation of the lake were published in 2011–2013, during which time urban construction was in a phase of rapid development, and thus, the rate of reduction of the lake was more stable. In addition, the intensity index can reﬂect the inﬂuences of human engineering activities on lakes. We found that the peak of −3.883 in 2003–2005 belongs to a burgeoning period of engineering activity in Wuhan. Meanwhile, the slow atrophy phase (2005–2013) exhibited a diminished rate of areal reduction with two brief falls in the lake area, which may have been related to the improved ecologicalFigure environment. 8. Distribution of selected large, key lakes throughout Wuhan. In 2015, We also conducted field trips to every lake to improve the results of comparisons between the image interpretations and the actual lake changes. Sustainability 2018, 10, 149 9 of 16

2003–2005. The variability of the lake areas was dramatic during the next 7 years. Between 2007 and 2013, the intensity index dropped to values between −1.848 and −1.126 in 2011–2013, proving that the variation amplitude was clear and that the changes in the lake areas were going to slow down year by year.

Intensity Index of Lake Evolution (%)

Figure 7. Total changes in the lake areas in the Wuhan urban area (1987–2013).

We studied the annual variation in the area of South Prince Lake as an example for the period between 1987 and 2013. South Prince Lake is located in the middle-west of the Wuhan area (Figure 8) Sustainabilityand is2018 adjacent, 10, 149 to the Yangtze River. According to the water surface extracted from the remote 10 of 16 sensing imagery, we can see the evolution process as presented in Figures 9 and 10.

Figure 8. Distribution of selected large, key lakes throughout Wuhan. In 2015, We also conducted Figure 8. Distribution of selected large, key lakes throughout Wuhan. In 2015, We also conducted ﬁeld field trips to every lake to improve the results of comparisons between the image interpretations and trips tothe every actual lake lake tochanges. improve the results of comparisons between the image interpretations and the actual lake changes. Sustainability 2018, 10, 149 10 of 16

Figure 9. Evident areal changes of South Prince Lake (1987–2013). The area at the southeast corner of Figure 9. Evident areal changes of South Prince Lake (1987–2013). The area at the southeast corner of the lake was clearly reduced, and field observations revealed that industrial construction had filled the lakethe was lake clearly in 2013. reduced, and ﬁeld observations revealed that industrial construction had ﬁlled the lake in 2013.

Figure 10. Change indicators for South Prince Lake (1987–2013).

As shown in Figure 10, the data for each time period were negative among the three indicators, indicating that the overall ecological environment of South Prince Lake was in a state of reduction. The amplitude of variation and the yearly rangeability show that the lake was in an accelerated stage of atrophy between 1987 and 2005 and in a deceleration phase between 2005 and 2013. During the two periods of 1987–1994 and 2011–2013, the yearly rangeability reached −0.46 km2 and −0.31 km2, which does not constitute a large difference. This was perhaps due to slow urban development in 1987–1994, during which time human engineering activity was not substantial. In addition, additional laws and regulations for the preservation of the lake were published in 2011–2013, during which time urban construction was in a phase of rapid development, and thus, the rate of reduction of the lake was more stable. In addition, the intensity index can reflect the influences of human engineering activities on lakes. We found that the peak of −3.883 in 2003–2005 belongs to a Sustainability 2018, 10, 149 10 of 16

Figure 9. Evident areal changes of South Prince Lake (1987–2013). The area at the southeast corner of Sustainabilitythe lake2018 was, 10 clearly, 149 reduced, and field observations revealed that industrial construction had filled11 of 16 the lake in 2013.

Sustainability 2018, 10, 149 11 of 16 Figure 10. Change indicators for South Prince Lake (1987–2013). burgeoning period Figureof engineering 10. Change indicatorsactivity in for Wuhan. South Prince Meanwhile, Lake (1987–2013). the slow atrophy phase (2005–2013) exhibited a diminished rate of areal reduction with two brief falls in the lake area, which As shown in Figure 10, the data for each time period were negative among the three indicators, may have been related to the improved ecological environment. indicatingExisting that images the overall have ecological been able environment to fully describe of South the changes Prince Lake in the was lakes, in a butstate we ofrequire reduction. the Existing images have been able to fully describe the changes in the lakes, but we require the support of data models to forecast future trends in their areal evolution. As mentioned in the previous Thesupport amplitude of data of variationmodels to and forecast the yearly future rangeability trends in showtheir thatareal the evolution. lake was inAs an mentioned accelerated in stage the chapter, our study applies a quadratic exponential smoothing model to predict the changes over the ofprevious atrophy chapter, between our 1987 study and applies 2005 and a quadratic in a decel exponentialeration phase smoothing between model 2005 to and predict 2013. the During changes the next 2 years in the Wuhan urban area. After repeated tests, we found that the average2 error will be2 twoover periods the next of 2 1987–1994 years in the and Wuhan 2011–2013, urban area.the yearly After repeatedrangeability tests, reached we found −0.46 that km the and average −0.31 error km , minimized when the smoothing parameter is set to 0.9. This is because both the water environment whichwill be does minimized not constitute when athe larg smoote difference.hing parameter This was is perhapsset to 0.9. due This to slowis because urban bothdevelopment the water in and human factors, which are greatly responsible for changes in the lakes, will continue to affect the 1987–1994,environment during and humanwhich factors,time human which engineeringare greatly activityresponsible was for not changes substantial. in the In lakes, addition, will lake areas, and the current rate of decrease in the lake area is closely associated with the ecological additionalcontinue to laws affect and the regulations lake areas, fo andr the the preservation current rate ofof thedecrease lake we in rethe published lake area inis closely2011–2013, associated during environment of the previous year. Therefore, the smoothing parameter will undoubtedly be relatively whichwith thetime ecological urban construction environment was of in thea phase previous of rapid year. development, Therefore, the and smoothing thus, the rateparameter of reduction will oflarge.undoubtedly the Takinglake was Sand be morerelatively Lake, stable. South large. In Prince Takingaddition, Lake, Sand the and Lake in Southtensity, South Lake index Prince as examples, can Lake, reflect and the theSouth modeling influences Lake results as examples,of andhuman the engineeringactualthe modeling values activities are results compared and on thelakes. in actual Figure We values 11found. are that compared the peak in Figure of −3.883 11. in 2003–2005 belongs to a

FigureFigure 11.11. ArealAreal changeschanges ofof SouthSouth Lake, Sand Lake, and South Prince Prince Lake Lake (1987–2013) (1987–2013) and and their their predictedpredicted values values (2014–2015).(2014–2015). TheThe areas of the lakes al alll decreased with with increasing increasing time, time, and and the the fitted ﬁtted valuesvalues were were closer closer to to the the actual actual values, values, indicating indicating that that the the predicted predicte valuesd values of moreof more recent recent phases phases will bewill more be more accurate. accurate.

By applying Formulas (8) and (9), the MAE and relative error of these three forecast models are 0.75By and applying 0.13, respectively. Formulas (8) In and consideration (9), the MAE andthat relativethe lake error areas of theseare usually three forecast relatively models large, are the 0.75 andprediction 0.13, respectively. accuracy of In the consideration test model is that acceptab the lakele. areasTherefore, are usually the modeling relatively results large, prove the prediction that the accuracyexponential of the smoothing test model model is acceptable. is suitable Therefore, for the sh theort-term modeling forecasting results proveof lake that area the changes. exponential As smoothingmentioned model earlier, is our suitable group for collected the short-term area data forecasting for 82 lakes of for lake the area period changes. 1987–2013; As mentioned consequently, earlier, ourwe groupcan obtain collected predictions area data of the for area 82 lakesof each for lake the for period 2015 1987–2013;by applying consequently, the exponential we smoothing can obtain predictionsmodel, the ofresults the area of ofwhich each are lake compiled for 2015 byin applyingTable 2, and the exponentialwe have completed smoothing field model, measurement the results ofwork which of all are lakes compiled area in in 2015. Table 2, and we have completed ﬁeld measurement work of all lakes area in 2015.Table 2 shows that the average error was 0.23 and that the correlation coefficient between the two data sets was 0.98. This indicates that the fitted data and the actual values have a high correlation, and therefore, the model forecasting ability is reliable. The largest predicted error was 2.281 km2 for lake 76, while lake 35 provided the smallest error of 0.002 km2. From Table 2, we can also conclude that the model predicted a lake area reduction of 18.494 km2 in 2015. The average error reached 0.23 with a correlation coefficient of 0.98, indicating that the model is reliable. The modeling results also show that in the next two years of 2013, the lake area in the Wuhan urban district will continue to slowly shrink, but the rate of decrease will have obviously slowed down relative to the previous ten years. The ambition of current lake conservation measures is to stop the shrinkage of the lakes by enhancing their supervision and enforcing current construction policies created by the governmental departments. However, many more incentives to protect the lakes should be introduced.

Sustainability 2018, 10, 149 12 of 16

Table 2. Lake area prediction results in the Wuhan urban area.

Lake No. 2015 (km2) Predicted Lake No. 2015 (km2) Predicted 1 0.155 0.135 42 0.132 0.024 2 0.381 0.316 43 0.296 0.243 3 3.340 2.940 44 0.723 0.654 4 3.356 3.165 45 1.900 1.842 5 1.266 1.013 46 1.594 1.443 6 0.238 0.222 47 0.101 0.081 7 0.105 0.096 48 0.978 0.799 8 1.537 1.511 49 5.753 5.641 9 1.168 1.064 50 0.701 0.642 10 0.605 0.554 51 0.250 0.224 11 0.191 0.177 52 3.479 3.074 12 17.020 16.913 53 0.960 0.938 13 0.576 0.549 54 0.897 0.756 14 0.283 0.199 55 0.436 0.396 15 0.280 0.267 56 0.424 0.411 16 0.070 0.059 57 29.520 28.461 17 0.074 0.064 58 1.146 1.078 18 0.090 0.073 59 2.819 2.433 19 30.586 29.462 60 0.684 0.465 20 11.668 10.446 61 0.043 0.035 21 0.110 0.076 62 0.033 0.025 22 2.479 2.046 63 7.714 7.456 23 0.314 0.233 64 0.570 0.421 24 5.538 5.113 65 2.801 2.465 25 0.087 0.084 66 0.241 0.213 26 2.071 1.997 67 1.126 1.018 27 7.419 6.944 68 1.679 1.412 28 3.378 3.294 69 14.874 14.657 29 0.318 0.279 70 2.108 2.011 30 6.789 6.513 71 0.934 0.796 31 0.203 0.179 72 0.800 0.649 32 0.233 0.193 73 3.880 3.155 33 0.334 0.244 74 11.187 10.842 34 0.556 0.397 75 27.531 27.166 35 0.073 0.071 76 42.780 40.499 36 9.457 9.046 77 3.262 3.154 37 0.202 0.183 78 28.586 26.744 38 2.072 1.974 79 2.497 2.018 39 1.773 1.465 80 0.029 0.027 40 2.088 1.974 81 1.550 1.510 41 0.206 0.188 82 8.594 8.211

Table2 shows that the average error was 0.23 and that the correlation coefficient between the two data sets was 0.98. This indicates that the fitted data and the actual values have a high correlation, and therefore, the model forecasting ability is reliable. The largest predicted error was 2.281 km2 for lake 76, while lake 35 provided the smallest error of 0.002 km2. From Table2, we can also conclude that the model predicted a lake area reduction of 18.494 km2 in 2015. The average error reached 0.23 with a correlation coefficient of 0.98, indicating that the model is reliable. The modeling results also show that in the next two years of 2013, the lake area in the Wuhan urban district will continue to slowly shrink, but the rate of decrease will have obviously slowed down relative to the previous ten years. The ambition of current lake conservation measures is to stop the shrinkage of the lakes by enhancing their supervision and enforcing current construction policies created by the governmental departments. However, many more incentives to protect the lakes should be introduced. Sustainability 2018, 10, 149 13 of 16

4.2. Lake Change Mapping Sustainability 2018, 10, 149 13 of 16 The predicted model can only reflect quantifiable changes in the lake areas. Therefore, to qualify theoverall lake area atrophy changes, of the we lakes rasterized was a common the shapes phenomenon of the lakes throughout over every the two research years and area. applied Through a change an investigation of the history of Wuhan development in recent years, the regions showing larger lake detection algorithm within ENVI using a classification method to obtain three different categories: lake reductions were located in areas with continuous urban engineering construction and rapid increase, lake decrease and no change. Finally, all of the layers were imported into ArcGIS, and we population growth, and thus, there was a heavy influence of human activity on those lakes, and the generatedtransformations a lake change of the lake mapping ecologic (Figureal environments 12). were tremendous.

Figure 12. Map of lake area changes in the Wuhan urban area. Figure 12. Map of lake area changes in the Wuhan urban area. We can conclude from the mapping and statistical results that the total lake area within the urbanFrom development the relative spatialzone tended distribution to decrease of the in mapped 1987–2013. lake These changes, changes many took shrinking place within lakes appearedthree in nearlystages: everythe lake partition area shrank of the gradually Wuhan development during 1987–2001, district, reduced which more indicates quickly that during an overall 2001–2005, atrophy of theand lakes gradually was decreased a common again phenomenon during 2005–2015. throughout According the research to relevant area. government Through historical an investigation data of thefor 1987–2001, history of the Wuhan primary development reason for the in reduction recent years, in the the lake regions area was showing the transformation larger lake of reductions many werelakes located into inponds areas to with acquire continuous arable land urban [42]. engineering Meanwhile, construction the lake area and shrank rapid populationmore quickly growth, in and2001–2005 thus, there as wasa consequence a heavy inﬂuence of the speed of human of urbanizat activityion, on which those lakes,led to andmore the land transformations for construction, of the and parts of the lake areas were filled. Finally, in 2005–2015, the shrinkage gradually decreased lake ecological environments were tremendous. because the Wuhan government strengthened the legislation for protecting the lake environments; We can conclude from the mapping and statistical results that the total lake area within the consequently, the ecological health and stability of the major lakes have recovered. urban development zone tended to decrease in 1987–2013. These changes took place within three stages:5. Conclusions the lake area shrank gradually during 1987–2001, reduced more quickly during 2001–2005, and gradually decreased again during 2005–2015. According to relevant government historical data for Our model was suitable for the prediction of water body variations, which could provide 1987–2001, the primary reason for the reduction in the lake area was the transformation of many lakes intuitive references for further environmental protection and management practices for lakes in the into ponds to acquire arable land [42]. Meanwhile, the lake area shrank more quickly in 2001–2005 Wuhan area. Based on a long-term series of remote sensing imagery in conjunction with field survey as adata consequence in the Wuhan of theurban speed area, of we urbanization, conducted a comprehensive which led to moreanalysis land of the for developmental construction, andtrends parts of theof the lake Wuhan areas urban were area ﬁlled. and Finally, some key in 2005–2015,lakes. Then, thewe shrinkageproduced an gradually easily understandable decreased because map of the Wuhanthe extent government of lake strengthenedarea changes thefor legislationWuhan that for is protecting suitable for the use lake atenvironments; local scales. Only consequently, a few theimportant ecological scenic health spots and exhibited stability ofsignificant the major growth lakes haveor did recovered. not change (e.g., East Lake), our results Sustainability 2018, 10, 149 14 of 16

5. Conclusions Our model was suitable for the prediction of water body variations, which could provide intuitive references for further environmental protection and management practices for lakes in the Wuhan area. Based on a long-term series of remote sensing imagery in conjunction with field survey data in the Wuhan urban area, we conducted a comprehensive analysis of the developmental trends of the Wuhan urban area and some key lakes. Then, we produced an easily understandable map of the extent of lake area changes for Wuhan that is suitable for use at local scales. Only a few important scenic spots exhibited significant growth or did not change (e.g., East Lake), our results showed that the environment and ecology were maintained in a relatively steady state; but in the urban of the Wuhan development area, big lakes (e.g., South Prince Lake) faced a danger of shrinking every year, because many real estate projects had been carried on around lakes, and some parts were filled directly for industrial use or fish farming; because of the excretion of water movement, we also found a more evident atrophy and shrinkage of the lakes in closer proximity to the coast of the Yangtze River (e.g., Sand Lake). Through an investigation of the history of Wuhan development in recent years, the regions showing larger lake reductions were located in areas with continuous urban engineering construction and rapid population growth, and thus, there was a heavy influence of human activity on those lakes, and the transformations of the lake ecological environments were tremendous. Furthermore, this research forecasted the areal changes in the lake surfaces by applying an exponential smoothing method. Regarding additional data sources for further research, further research will carry more comprehensive investigation and meaningful models on environmental and ecological changes in lakes around Wuhan. Because other external factors that affect the shrinkage of lakes must be observed and recorded so we can obtain a more robust and accurate model.

Acknowledgments: This research was funded by the National High-Tech Research and Development Program of China (2012AA121303), and by the Engineering Research Center of Geospatial Information and Digital Technology (SIDT20170201), NASG. We are also grateful to the help of Wuhan surveying and mapping institute. Author Contributions: G.H.D. and R.Q.N. conceived the experiments; G.H.D. developed the model code and processed the data and developed the graphics with assistance from R.Q.N. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. Zha, Y.; Gao, J.; Ni, S. Use of normalized difference built-up index in automatically mapping urban areas from TM imagery. Int. J. Remote Sens. 2003, 24, 583–594. [CrossRef] 2. Zhu, Z.; Wang, S.; Woodcock, C.E. Improvement and expansion of the Fmask algorithm: Cloud, cloud shadow, and snow detection for Landsats 4–7, 8, and Sentinel 2 images. Remote Sens. Environ. 2015, 159, 269–277. [CrossRef] 3. Jackson, L.E. The relationship of urban design to human health and condition. Landsc. Urban Plan. 2003, 64, 191–200. [CrossRef] 4. Ehrenfeld, J.G. Evaluating wetlands within an urban context. Ecol. Eng. 2000, 15, 253–265. [CrossRef] 5. Feyisa, G.L.; Meilby, H.; Fensholt, R.; Proud, S.R. Automated Water Extraction Index: A new technique for surface water mapping using Landsat imagery. Remote Sens. Environ. 2014, 140, 23–35. [CrossRef] 6. Tulbure, M.G.; Broich, M.; Stehman, S.V.; Kommareddy, A. Surface water extent dynamics from three decades of seasonally continuous Landsat time series at subcontinental scale in a semi-arid region. Remote Sens. Environ. 2016, 178, 142–157. [CrossRef] 7. Bai, S.B.; Wang, J.; Zhou, P.-G.; Hou, S.S.; Lü, G.-N.; Xu, S.N. GIS-based logistic regression for landslide susceptibility mapping of the zhongxian segment in the Three Gorges area, China. Geomorphology 2010, 115, 23–31. [CrossRef] 8. Fourniadis, I.G.; Liu, J.G.; Mason, P.J. Landslide hazard assessment in the Three Gorges area, China, using ASTER imagery: Wushan–Badong. Geomorphology 2007, 84, 126–144. [CrossRef] Sustainability 2018, 10, 149 15 of 16

9. Turcotte, R.; Fortin, J.P.; Rousseau, A.N.; Massicotte, S.; Villeneuve, J.P. Determination of the drainage structure of a watershed using a digital elevation model and a digital river and lake network. J. Hydrol. 2001, 240, 225–242. [CrossRef] 10. Mueller, N.; Lewis, A.; Roberts, D.; Ring, S.; Melrose, R.; Sixsmith, J.; Lymburner, L.; McIntyre, A.; Tan, P.; Curnow, S.; et al. Water observations from space: Mapping surface water from 25 years of Landsat imagery across Australia. Remote Sens. Environ. 2016, 174, 341–352. [CrossRef] 11. Xu, H.Q. Modification of normalised difference water index (NDWI) to enhance open water features in remotely sensed imagery. Int. J. Remote Sens. 2006, 27, 3025–3033. [CrossRef] 12. Fisher, A.; Flood, N.; Danaher, T. Comparing Landsat water index methods for automated water classification in eastern Australia. Remote Sens. Environ. 2016, 175, 167–182. [CrossRef] 13. Mcfeeters, S.K. The use of Normalized Difference Water Index (NDWI) in the delineation of open water features. Int. J. Remote Sens. 1996, 17, 1425–1432. [CrossRef] 14. Xu, H.Q. Dynamic monitoring of land-use/land-cover change and urban expansion in Shenzhen using Landsat imagery from 1988 to 2015. Int. J. Remote Sens. 2006, 38, 5388–5407. 15. Wilson, E.H.; Sader, S.A. Detection of forest harvest type using multiple dates of Landsat TM imagery. Remote Sens. Environ. 2002, 80, 385–396. [CrossRef] 16. Eldhuset, K. Combination of stereo SAR and InSAR for DEM generation using TanDEM-X spotlight data. Int. J. Remote Sens. 2017, 38, 4362–4378. [CrossRef] 17. Chibani, Y.; Houacine, A. The Joint use of HIS Transform and redundant wavelet decomposition for fusing multispectral and panchromatic images. Int. J. Remote Sens. 2002, 23, 3821–3833. [CrossRef] 18. FujitK, A.; Suzuki, R.; Nuimura, T.; Sakai, A. Performance of ASTER and SRTM DEMs, and their potential for assessing glacial lakes in the Lunana region, Bhutan Himalaya. J. Glaciol. 2008, 54, 220–228. [CrossRef] 19. Kiema, J.B.K. Texture analysis and data fusion in the extraction of topographic objects from satellite imagery. Int. J. Remote Sens. 2002, 23, 767–776. [CrossRef] 20. Rundquist, D.C.; Lawson, M.P.; Queen, L.P.; Cerveny, R.S. The relationship between summer-season rainfall events and lake-surface Area. Water Resour. Bull. 1987, 23, 493–508. [CrossRef] 21. Gill, T.; Johansen, K.; Phinn, S.; Trevithick, R.; Scarth, P.; Armston, J. A method for mapping Australian woody vegetation cover by linking continental-scale field data and long-term Landsat time series. Int. J. Remote Sens. 2017, 38, 679–705. [CrossRef] 22. Du, Z.Q.; Li, W.B.; Zhou, D.B.; Tian, L.Q.; Ling, F.; Wang, H.L.; Gui, Y.M.; Sun, B.Y. Analysis of Landsat-8 OLI imagery for land surface water mapping. Remote Sens. Lett. 2014, 5, 672–681. [CrossRef] 23. Ouma, Y.O.; Tateishi, R. A water index for rapid mapping of shore line changes of five east African rift valley lakes: An empirical analysis using Landsat TM and ETM + data. Int. J. Remote Sens. 2006, 27, 3153–3181. [CrossRef] 24. Lacoste, C.; Descombes, X.; Zerubia, J. Unsupervised line network extraction in remote sensing using a polyline process. Pattern Recognit. 2010, 43, 1631–1641. [CrossRef] 25. Zitov, B.; Flusser, J. Image registration methods: A survey. Image Vis. Comput. 2003, 21, 977–1000. [CrossRef] 26. Pozzi, F.; Matteo, T.D.; Aste, T. Exponential smoothing weighted correlations. Eur. Phys. J. B 2012, 85, 175. [CrossRef] 27. Monfared, M.A.S.; Ghandali, R.; Esmaeili, M. A new adaptive exponential smoothing method for non-stationary time series with level shifts. J. Ind. Eng. Int. 2014, 10, 209–216. [CrossRef] 28. Pee, E.Y.; Royset, J.O. On solving large-scale finite minimax problems using exponential smoothing. J. Optim. Theory Appl. 2011, 148, 390–421. [CrossRef] 29. Zhou, S.; Cui, J.; Ye, F.; Liu, H.; Zhu, Q. New smoothing SVM algorithm with tight error bound and efficient reduced techniques. Comput. Optim. Appl. 2013, 56, 599–617. [CrossRef] 30. Tangxun Lake, the Largest Lake in Wuhan City, Completed the Rectification of the Area Occupation. Available online: http://news.xinhuanet.com/2017-08/06/c_1121439067.htm (accessed on 6 August 2017). 31. Roy, D.P.; Wulder, M.; Loveland, T.; Woodcock, C.; Allen, R.; Anderson, M.; Helder, D.; Irons, J.; Johnson, D.; Kennedy, R.; et al. Landsat-8: Science and product vision for terrestrial global change research. Remote Sens. Environ. 2014, 145, 154–172. [CrossRef] 32. Lu, D.; Weng, Q.A. Survey of image classification methods and techniques for improving classification performance. Int. J. Remote Sens. 2007, 28, 823–870. [CrossRef] Sustainability 2018, 10, 149 16 of 16

33. Gao, B.C. NDWI—A normalized difference water index for remote sensing of vegetation liquid water from space. Remote Sens. Environ. 1996, 58, 257–266. [CrossRef] 34. Taylor, J.W. Exponential smoothing with a damped multiplicative trend. Int. J. Forecast. 2003, 19, 715–725. [CrossRef] 35. Corberan-Vallet, A.; Bermudez, J.D.; Vercher, E. Forecasting correlated time series with exponential smoothing models. Int. J. Forecast. 2011, 27, 252–265. [CrossRef] 36. Gardner, E.S., Jr.; Mckenzie, E. Note—Seasonal exponential smoothing with damped trends. Manag. Sci. 1989, 35, 372–376. [CrossRef] 37. Olaofe, Z.O. Wind energy predictions of small-scale turbine output using exponential smoothing and feed-forward neural network. Int. J. Energy Eng. 2015, 5, 28–42. [CrossRef] 38. Spelta, C.; Previdi, F.; Savaresi, S.M.; Bolzern, P.; Cutini, M.; Bisaglia, C.; Bertinotti, S.A. Performance analysis of semi-active suspensions with control of variable damping and stiffness. Veh. Syst. Dyn. 2011, 49, 237–256. [CrossRef] 39. Hyndman, R.J.; Koehler, A.B.; Snyder, R.D.; Grose, S. A state space framework for automatic forecasting using exponential smoothing methods. Int. J. Forecast. 2002, 18, 439–454. [CrossRef] 40. Hyndman, R.J.; Khandakar, Y. Automatic Time Series Forecasting: The forecast Package for R. J. Stat. Softw. 2008, 27, 1–22. [CrossRef] 41. Dong, Z.; Yang, D.; Reindl, T.; Walsh, W.M. Short-term solar irradiance forecasting using exponential smoothing state space model. Energy 2013, 55, 1104–1113. [CrossRef] 42. Compiling committee of Hubei water resources development history. In Hubei Water Resources Development History; China Water Power Press: Beijing, China, 2000; pp. 254–256. (In Chinese)

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).