Land Cover Change And Its Climate Effect In The Three Gorges Reservoir Area

The transformation of land cover is very important to the research of global environment changes and sustainable growth.The Reservoir Region of Changjiang Three Gorges(RRTG),lying at the core of Changjiang Economic Zone,has become a hot spot for researches related to land covers and climate effects,as the RRTG is a typical example for its unique geographical environment and difficult land utilization.Most of the studies focus on single factor(land cover or body of water),few related to the expansion of land cover transformation and its climate effects.With the implement of the Development Plan of Changjiang Economic Zone,the RRTG should stick to the principle of “ecological and green development”,studies about spatial pattern,dynamic mechanism and climate effects are needed.We need to pay attention to the driving force of land use policy and reveal the relationship between land cover transformation and its climate effects.Social,economic and environmental benefits are three inseparable parts when perusing the goal of scientific land resource utilization.Our study is based on multi-temporal remote sensing images,water system and DEM.According to geographic characteristics and inundation time,a classification system including the non-flooded area and flooded area is built to compare the macroscopic spatial and temporal patterns in different stages(1997-2003,2003-2006 and 2006-2009)and analyze the transformation of mountain-valley-water interfaces and the terrain changes.Based on four stages' data(1990,2000,2010 and 2015),we set up a multiple-model system and take several main factors evaluate the driving force of land cover transformation: Using ESDA to analyze the driving force and its differences in spatial patterns;using Binary Logistic model to study the relationships between driving factors and land covers which based on the statistics of 2000.The model of CLUE-S is utilized to imitate the land cover in 2000.we imitate the land cover transformation in 2020 and in 2030 respectively according to natural development,food security,immigrations and environment protection.Taking daily statistics from 1951 to 2009 in 34 national climate sites,we get to know the changes of climate in the same period thus analyze its variation trend and disastrous weather.Using the revised P-M model,we evaluate the natural responses toward land cover transformation,like potential evapotranspiration and the moisture index(which is the ratio between precipitation and evapotranspiration).All these studies are aimed at giving scientific explanation for the land cover transformation of reservoir region and can be a firm support for land utilization.The results shows that:1.In different stages,the land cover within project region undergoes different and multiplied transformation.The seasonal green covers and evergreen covers constitute the main permeable landscape structure,which take up 60% of the whole area.The extension and scale of land cover transformation is different in different stages.The areas of transformation in 1997-2003,2003-2006 and 2006-2009 are 1313.60,683.30 and 771.14 km~2,respectively.The transformation is overall consistent: the change between seasonal and evergreen cover is mutual,grey covers gradually occupy seasonal green covers,the rising water submerge the green and blue cover.But the land cover transformation at partial region are significantly different.In the first two stages(1997-2003,2003-2006),the transformation of mountain-mountain is dominant,whose areas are 861.78 and 467.21 square kilometers,respectively,which takes up 65% of the total transformation area.In the last stage(2006-2009),the transformation is mainly between mountains(with an area of 443.94 km~2)and valleys when homogeneous land covers are converted into different type in different stages.The distribution of land cover transformation is consisted with natural development and human activities considering slope and altitude: During 1997 to 2003,areas below 500 meters are the transformation-prone region.From 2003 to 2006,the transformation-prone region are areas of 175~ 500 meters,500 ~1000 meters and 1500~2000 meters.From 2006 to 2009,they are areas of 175~500 meters and 1500~2000 meters.2.The driving force evaluation of recent 25 years shows that the city zone is much more transformed than other places.From 1990 to 2015,the comprehensive land cover transformation value of RRTG is increasing,when the dominant driving factors are human activities,population and transportation.The regions near the city zone of Chongqing and the places next to Yichang are urban-driven type.The places in northeast Chongqing,like Badong,Zigui and Xingshan of Hubei,are less driven by natural and economic forces,which are grassland-water-driven type.The other regions are multi-factor-driven type.In 1990,2000,2010 and 2015,the Moran's I value is 0.487,0.530,0.345 and 0.373(with Z>1.96),which suggests that the integrated spatial correlation degree is strong but decreasing,while the partial spatial correlation degree is based on regional social-economic conditions.Based on the difference of regional land cover transformation force,the strategy of land resource optimization can be conducted as follow: city zone of Chongqing and Yichang should take comprehensive utilization measures toward land resources;For other regions of RRTG,seeking for economic opportunities should base on environment protection.3.The ROC value of paddy fields,dry lands,grasslands,woodlands,construction land and waters are all above 0.8,which means the driving factors in Auto-logistic model that based on neighborhood abundance have a strong explanation ability,thus the probability distribution can be reckoned according to the model.In 2010,the Kappa coefficients for paddy fields,dry lands,woodlands,grasslands,construction lands and waters are 0.90,0.92,0.97,0.84,0.85 and 0.77 respectively.Generally,woodlands are best imitated while waters are relatively weak.The Kappa test shows that imitation of waters can meet the need for prediction.The imitation shows the competence of different land covers and their influence on food security,immigration and environment protection(including the transformation from paddy fields to dry lands,the occupation of cultivated lands or grasslands by construction land,the cultivation of woodlands,the transformation of cultivated lands to woodlands)and suggests the balance in the optimization process is very important.For natural growth between 2010 and 2030,the area transformed from paddy fields to dry lands is estimated to be 1411.25 km~2,which can be seen in all the counties of Chongqing,especially in the gentle slopes.As for food security,the area of cultivated lands is shrinking while most of the transformation of paddy field to dry lands is from 2010 to 2020(with an area of 666 km~2 and less than the half of the amount of natural growth).For the immigration resettlement,the construction areas are in rapid expansion and influenced more by immigration than natural development and food security,while most of the new land are from paddy fields,grassland and dry lands(with a total area of 2146.25 km~2).For environment protection,the protection and recovery for grassland and woodland of high ecological value are the most important goal,from 2010 to 2020,the area of returning the farmland to forest can reach 2251 km~2.4.The construction of reservoir may enlarge the meteorological fluctuation of RRTG.The annual average temperatures,annually highest temperature and the annually lowest temperature are increasing while the annual precipitation,annual sunshine duration and annual wind speed are in decrease.Though the region undergoes an abrupt change of climate,the construction of reservoir is far from being the main driving force for the change.Most of the climate changes take place before the filling.For the changes of meteorological factors from 1997 to 2003,the variation of average temperature difference(-0.65~8.64 ?)and relative hygrometric deficit(-3.86%~4.55%)is comparatively low while those of precipitation difference(-59.06~624.4 mm)and sunshine duration(-186.35~305.49 h)are comparatively high.From 2003 to 2006,the average precipitation and relative hygrometric deficit are negative within the reservoir region while the differences in temperature value and sunshine duration in Chongqing are in counter-parallelism with those in Hubei.From 2006 to 2009,the temperature and sunshine duration are decreasing annually while the precipitation is increasing.The annual relative humidity is increasing in general while decreases at some points.About the changes related to meteorological factors in the main land cover of RRTG,the main types undergo an increase in temperature from 1997 to 2003,which is more stable than other meteorological factors.The 2124 type(from seasonal green cover to grey cover)has an average temperature of 18.05 and 18.43? in 1997 and 2003.The annual precipitation is increasing obviously and the relative humidity is polarized.The changes of sunshine duration are in reverse situation with annual relative humidity.From 2003 to 2006,the average temperatures of main types are increasing and those of 2124 type is in high increase in 2006(19.04 ?).In this period,the fluctuation of precipitation is more undulated than other factors.The curve of relative humidity average value shows a “Contrary W” shape,but the variation of annual sunshine hours shows a “W” type.Both peak and valley values are in the 2221 type(from evergreen cover to seasonal green cover).In the period of 2006 to 2009,the precipitation for all of the main types is increasing while temperature and sunshine duration are in decrease.5.From 1997 to 2009,the average of annual evapotranspiration was firstly increasing but then turned into decrease and reached the peak in 2006(976.71 mm).The annual moisture index is in reversion with annual potential evapotranspiration.In 1997,the lower reaches have more evapotranspiration than the upper reaches,while in 2003 the regions with higher evapotranspiration values moves westwards obviously.In 2006,the evapotranspiration value is increasing while in 2009 the area starts to shrink.The annual moisture index is only different in some points.From 1997 to 2009,different areas in the bottom part,middle part and head part of RRTG with different land covers have difference in potential evapotranspiration and moisture index.In the bottom part,the evapotranspiration is bigger than two other regions,whose average values are between [918.89,944.74].In the middle part,all types of land cover have a bigger evapotranspiration than in the head part.Compared to the other 2 regions,the head part experiences more fluctuated changes in 2006 for the value of evapotranspiration in both seasonal and eternal submerge areas is above 1000 mm.About the moisture index of land cover,the bottom part is increasing in annual average moisture index in 2003 while goes through little changes in other years.In 2003,the value of middle part is between 1.2 and 1.3,which drops obviously compared to the value of bottom part.From 2003 and 2009,the value of head parts increases obviously than that of middle part,especially in 2003 when the transformation from seasonal submerge area to grey land cover in submerge area undergoes an increase from 1.3 and 1.71.In total,the changes of annual potential evapotranspiration undergo both increase and decrease,while in the same period the moisture index is always increasing.From 2003 to 2006,the potential evapotranspiration is in increase while the moisture index is stable.From 2006 to 2009 the evapotranspiration is decreasing while moisture index is in increase.Through the studies on the transformation of land cover in RRTG and its climate effect,we find out the process and trajectory of land cover transformation in the region.We identify the driving factors in the transformation and imitate the possible situation in the future.The variation of meteorological factor and relevant response in moisture index are also studied.Our findings can enrich people's thought about the transformation of land cover in the reservoir region,which can be a scientific explanation and reference to land policies in the regions and a good example for the regions under the influence of great projects.