郭春霞
, 诸云强, 孙伟
GUO Chunxia, ZHU Yunqiang, SUN Wei
通讯作者:
收稿日期: 2015-04-13
修回日期: 2015-07-8
网络出版日期: 2015-09-15
版权声明: 2015 《地理研究》编辑部 《地理研究》编辑部
基金资助:
作者简介:
作者简介：郭春霞（1991- ）,女,山西大同人,硕士,研究方向为地理信息系统应用。 E-mail: guochunxia1991@163.com
展开
摘要
不同时间尺度、季节的气温数据表现出不同的空间平稳特征。为探讨分析空间平稳性对气温插值的影响规律,采用趋势线法对气温数据进行空间平稳性探索,并对比分析不同空间平稳性条件下,普通线性回归、普通克里格、回归克里格的气温插值精度及插值结果的空间分布特点。结果显示：冬季日均、月均气温与年均气温呈现空间非平稳,插值精度随时间序列的增长而提高,随着气温数据逐渐趋于稳定,精度提高的幅度逐渐下降;夏季日均、月均气温呈现空间平稳,随时间序列的增长,插值精度的提高并不显著;夏季日均气温各插值方法的插值精度普遍高于冬季日均气温。与普通克里格相比,回归克里格能有效提高空间非平稳数据的插值精度。时间序列的增长削弱了不同插值算法之间的插值精度差异和插值结果空间分布差异。
关键词：
Abstract
Spatial stationary is a hypothesis for most geo-statistical processes. In order to explore the influence of spatial non-stationary on air temperature interpolation performance, a set of air temperature data in different time scales (including daily average air temperature, monthly average air temperature of January and July and annual average air temperature in 2010) are used. First of all, stepwise regression analysis is adopted to select the most important regression parameters for each data set of temperature. Then trend line analysis is used to estimate whether the air temperature data meets the assumption of spatial stationary. Finally, ten-cross validation is carried out by using the interpolation methods of ordinary linear regression, ordinary kriging, and regression kriging respectively. According to the results, the conclusion can be summarized as follows: 1) Daily average, monthly average air temperature of January, and annual average air temperature data present spatial non-stationary characteristic and an obvious change trend in the north-south direction; while daily average and monthly average air temperature of July are spatial stationary. Interpolation accuracy of daily average air temperature in July, which is spatial stationary, is significantly higher than that in January, which is spatial non-stationary. 2) In general, all the three interpolation methods obtain the best prediction results on annual dataset, then monthly datasets, worst on daily datasets, because the spatial structure of the daily air temperature dataset is more non-stationary than that of the monthly and annual temperature datasets. In terms of time series, the interpolation error reduces with the decrease of the degree of reduction. 3) Regression kriging achieves higher interpolation accuracy on each dataset in general than ordinary kriging, furthermore the improvement of interpolation accuracy achieved by regression kriging is more obvious on non-stationary datasets than on stationary datasets. 4) Distribution of air temperature interpolated by various techniques presents significantly difference in daily time scale, but in monthly and annual scales, there is no significant difference. Values in long time series, which are the means of values in short time series, weaken the occurrence probability of extreme values. Thus the distribution ranges of air temperature in January and July decrease, compared with daily average air temperature in corresponding seasons.
Keywords：
气温空间插值是指根据有限的气象监测数据估算未知点处的气温,是离散站点气温规则格网化的重要手段之一。气温空间插值方法的对比研究[1-5]、区域应用[6,7]、方法优化[8-10],一直是空间插值领域的研究热点。彭彬等通过交叉验证对不同空间插值方法插值的江苏省2003年月均气温与年均气温的插值精度进行了对比,发现普通克里格法的插值精度明显优于反距离加权法和张力样条插值法,而协同克里格法由于在插值过程中考虑了多个显著影响气温空间分布的协变量,插值精度一般优于普通克里格[11]。杨凤海等采用协同克里格法对黑龙江省旬平均气温进行空间插值,在插值结果的基础上对黑龙江省近10年气温的时空变异特征进行了分析[6]。潘耀忠等提出的基于DEM智能搜索距离插值方法,与反距离加权法相比,插值精度提高了一倍以上,而且可以获得高空间分辨率的网格气温数据[12]。朱会义等讨论了自然地理要素进行空间插值时需要考虑的多个问题,由于时间尺度的不同,地理要素呈现出的空间变异程度、规律性差异显著[13]。
已有研究主要是针对某一地区某一特定时间尺度的气温数据进行插值算法的对比[14-17],但对于不同时间尺度、季节气温数据的空间平稳特征对插值精度及结果的影响还需要进一步研究讨论。选择全国2010年1月1日（冬季,下同）、7月1日（夏季,下同）的日平均气温,1月、7月的月平均气温和2010年的年平均气温作为源数据,讨论各个时间尺度的空间平稳特性及其对普通线性回归（ordinary linear regression, OLR）、普通克里格（ordinary kriging, OK）、回归克里格（regression kriging, RK）插值精度及插值结果空间分布特点的影响。
2010年气温数据来源于中国气象科学数据共享服务网（http://cdc.cma.gov.cn）,气象站点分布如图1所示。数据集包含838个气象站点数据,剔除坐标重复及高程异常的数据记录,得到插值的791条数据记录。全国1 km分辨率的DEM数据集与1
图1 全国气象台站分布图
Fig. 1 The spatial distribution of meteorological stations in China
图1显示中国气象站点集中分布在东部地区,西部地区气象站点分布稀疏,青藏高原海拔超过5000 m的地区,几乎没有气象站点分布。
为揭示不同时间尺度、不同季节气温监测数据的集中趋势、离散程度以及分布规律,选择平均值、中位数、最大值、最小值、标准差、峰度系数、偏度系数7个统计量（表1）对数据进行初步分析和判断。
表1 不同时间尺度、季节气温描述性统计量
Tab. 1 Descriptive statistics of air temperature in different time scales and seasons
| 统计量 | 平均气温（℃） | ||||
|---|---|---|---|---|---|
| 2010.01.01 | 2010.07.01 | 2010.01 | 2010.07 | 2010 | |
| 平均值 | -2.18 | 24.59 | -1.51 | 24.50 | 11.76 |
| 中位数 | 0.70 | 24.80 | -1.10 | 25.70 | 12.50 |
| 最小值 | -41.80 | 5.10 | -29.40 | 7.70 | -4.60 |
| 最大值 | 22.50 | 33.60 | 20.90 | 33.30 | 25.50 |
| 标准差 | 11.90 | 5.59 | 10.27 | 4.75 | 6.4 |
| 峰度系数 | 0.32 | 0.14 | -0.65 | 0.94 | -0.82 |
| 偏度系数 | -0.90 | -0.72 | -0.31 | -1.16 | -0.26 |
表1显示,所有791个气象台站中,2010年1月1日最低气温为-41.8 ℃,最高气温为22.5 ℃;2010年1月最低气温为-29.4 ℃,最高气温为20.9 ℃。2010年7月1日最低气温为5.1 ℃,最高气温为33.6 ℃;2010年7月最低气温为7.7 ℃,最高气温为33.3 ℃。数据显示,月平均最高气温低于日平均最高气温,月平均最低气温高于日平均最低气温。时间序列的增长会不同程度地平滑掉监测数据中的极值信息。冬季和夏季月均气温的标准差都低于日均气温,长时间序列平均削弱了短时间序列数据中的极值,使得所有站点数据更加集中。
结合已有研究,选择经度、纬度、高程、坡度、坡向和地表粗糙度6个地形因子,讨论其与日、月、年三个时间尺度,冬、夏两个季节五个气温数据集的相关程度,在相关分析基础上,采用逐步回归筛选出对各个回归模型具有显著贡献的回归参数。相关、逐步回归结果表明：2010年1月1日平均气温的回归参数为经度、纬度、高程和坡度;7月1日平均气温的回归参数为经度、纬度和高程;1月平均气温的回归参数为纬度、高程、坡度和地表粗糙度;7月平均气温的回归参数为经度、纬度和高程;2010年平均气温的回归参数为经度、纬度、高程和坡度。全局拟合优度
冬季日均气温与月均气温的回归参数不同,且月均气温的拟合精度高于日均气温;夏季日均气温与月均气温的回归参数相同,但时间序列的增长并没有明显提高模型的拟合精度;年均气温的回归参数与冬季日均气温相同,拟合精度高于日均气温却低于月均气温。
主要采用趋势线法分析讨论全国气温数据的空间平稳特点。趋势线法[18]是将气温站点数据分别投影到东西和南北两个方向上,通过两个方向的二维散点图发现数据是否空间平稳,以及在不满足空间平稳假设下的趋势走向。各时间尺度不同季节气温数据的趋势线如图2所示。
图2 不同时间尺度、季节气温东西和南北趋势图
Fig. 2 Trend lines of air temperature in different time scales and seasons in the east-west and south-north directions
根据图2a~图2d,可以发现冬季气温在南北方向上存在显著的气温随纬度增加而下降的趋势,而且月均气温的趋势较日均气温更加显著,因此冬季气温并不满足空间平稳假设。根据图2e~图2h,可以发现气温站点在任意方向上的分布都是随机的,并不存在倾向于某个方向的趋势,因此夏季气温是空间平稳的。根据图2i~图2j,可以发现年均气温同样在南北方向上存在显著的趋势,趋势走向与冬季日均气温和月均气温相同,因此年均气温也不满足空间平稳假设。
通过十折交叉验证法对各个插值方法进行精度验证,将观测数据集随机分为10份,每次验证抽取其中的9份数据作为建模集,剩余1份数据作为验证集,分别计算均方根误差RMSE（式1）和平均绝对误差MAE（式2）两个指标,最后求取10次验证结果的平均RMSE和MAE。
式中：
表2 冬夏两季日均气温插值精度
Tab. 2 Interpolation accuracy of daily average air temperature in winter and summer
| 插值方法 | 1月1日平均气温(℃) | 7月1日平均气温(℃) | |||
|---|---|---|---|---|---|
| RMSE | MAE | RMSE | MAE | ||
| 普通线性回归 | 4.454 | 3.470 | 2.790 | 2.192 | |
| 普通克里格 | 4.032 | 3.064 | 2.537 | 1.752 | |
| 回归克里格 | 2.559 | 1.816 | 2.061 | 1.388 | |
表3 冬夏两季月均气温插值精度
Tab. 3 Interpolation accuracy of monthly average air temperature in winter and summer
| 插值方法 | 1月平均气温(℃) | 7月平均气温(℃) | |||
|---|---|---|---|---|---|
| RMSE | MAE | RMSE | MAE | ||
| 普通线性回归 | 2.929 | 2.139 | 2.332 | 1.706 | |
| 普通克里格 | 2.490 | 1.809 | 2.533 | 1.731 | |
| 回归克里格 | 1.750 | 1.195 | 1.964 | 1.265 | |
表4 年均气温插值精度
Tab. 4 Interpolation accuracy of annual average air temperature
| 插值方法 | 2010年平均气温(℃) | |
|---|---|---|
| RMSE | MAE | |
| 普通线性回归 | 1.981 | 1.408 |
| 普通克里格 | 2.361 | 1.625 |
| 回归克里格 | 1.784 | 1.133 |
表2显示,三种插值方法7月1日平均气温的插值精度普遍高于1月1日平均气温。1月1日回归克里格的插值误差较普通克里格RMSE降低了36.5%,MAE降低了40.7%;7月1日RMSE降低了18.8%,MAE降低了20.8%。表3显示,1月平均气温回归克里格的插值误差较普通克里格RMSE降低了29.7%,MAE降低了33.9%;7月RMSE降低了22.5%,MAE降低了26.9%。表4显示,年平均气温回归克里格较普通克里格的插值误差,RMSE降低了24.4%,MAE降低了30.3%。由于回归克里格有效剔除了空间非平稳数据存在的趋势,因此插值精度较普通克里格大幅度提高;而对于并不存在显著趋势的空间平稳数据,插值精度的提高幅度不如空间非平稳数据显著。
综合分析表2~表4,冬季普通线性回归RMSE分别降低了34.2%和32.4%;MAE分别降低了38.4%和34.2%。普通克里格的插值误差从日尺度到月尺度再到年尺度,RMSE分别降低了38.2%和5.2%;MAE分别降低了41%和10.2%。回归克里格RMSE分别降低了31.6%和1.9%;MAE分别降低了34.2%和5.2%。夏季普通线性回归RMSE分别降低了16.4%和15.1%;MAE分别降低了22.2%和17.5%。普通克里格的插值误差从日尺度到月尺度再到年尺度,RMSE分别降低了0.2%和6.8%;MAE分别降低了1.2%和6.1%。回归克里格RMSE分别降低了4.7%和9.2%;MAE分别降低了8.9%和10.4%。冬季随着时间序列的不断增长,普通克里格与回归克里格插值误差的下降幅度显著变小,而普通线性回归的下降幅度基本持平。夏季随着时间序列的增长,插值误差的下降幅度没有一定的规律性。
为探讨不同时间尺度、季节气温采用三种不同插值方法所得插值结果的空间分布特点及结构差异,利用统计分析软件R分别对日、月、年三个时间尺度,冬、夏两个季节的气温进行空间插值及可视化表达,气温分级分布如图3~图7所示。
图3 全国冬季日均气温分级分布图
Fig. 3 The spatial distribution of classified daily average air temperature in winter in China
图4 全国冬季月均气温分级分布图
Fig. 4 The spatial distribution of classified monthly average air temperature in winter in China
图5 全国夏季日均气温分级分布图
Fig. 5 The spatial distribution of classified daily average air temperature in summer in China
图6 全国夏季月均气温分级分布图
Fig. 6 The spatial distribution of classified monthly average air temperature in summer in China
图7 全国年均气温分级分布图
Fig. 7 The spatial distribution of classified annual average air temperature in China
图3为全国1月1日平均气温三种插值结果的空间分布情况。从气温整体分布趋势及特点而言,主要表现为局部差异,如青藏高原。三种插值结果的极值显示：与气象站点的最低记录气温相比,普通线性回归偏高,普通克里格偏低,回归克里格略高,最接近于最低记录气温;与气象站点的最高记录气温相比,普通线性回归略低,而普通克里格与回归克里格二者相近,高于最高记录气温。
图4为全国1月平均气温三种插值结果的空间分布情况。三种插值结果的气温分布同样表现为局域差异,与日均气温相比,全国自北向南气温过渡较为平缓,最低气温分布范围变小,主要表现在青藏高原与东北北部地区,普通线性回归与回归克里格最为显著。插值结果的极值显示：普通克里格的最低气温远低于站点最低记录气温,最高气温远高于站点最高记录气温;回归克里格最低气温低于普通线性回归,与站点最低记录气温基本一致,最高气温与普通线性回归基本一致且高于站点最高记录气温。
图5显示,全国7月1日平均气温整体上自西北至东南逐渐升高。三种插值结果的气温分布差异主要表现在内蒙古北部、新疆、青藏高原及东北地区。插值结果极值显示：普通克里格的最低、最高气温分别略高于气象站点最低、最高记录气温,普通线性回归与回归克里格的最低气温低于站点最低记录气温,最高气温略低于站点最高记录气温。
图6显示,7月平均气温的三种插值结果主要在东北、华北和华南地区呈现差异。极值特点类似于7月1日,然而普通克里格的高气温低于站点最高记录气温。高于28.75 ℃的极高气温带较7月1日显著缩小,山西周边一带气温有所提高,东北北部气温有所下降。
图7显示,普通线性回归与回归克里格的插值结果气温分布呈现局部细微差异。插值结果极值显示：所有插值结果最低气温低于气象站点最低记录气温,最高气温高于气象站点最高记录气温。普通克里格到普通线性回归再到回归克里格,值域范围呈现小幅度缩小。
为探讨不同时间尺度、季节气温数据的空间平稳特征及其对不同插值方法插值精度及插值结果空间分布的影响,分别选择了冬季日均气温和月均气温（2010年1月1日、2010年1月）,夏季日均气温和月均气温（2010年7月1日、2010年7月）,以及年均气温（2010年）作为数据源。在相关性分析基础上,通过逐步回归选取不同时间尺度、季节气温的回归参数,并通过趋势线法分析了各自的空间平稳特征。分别采用普通克里格、普通线性回归以及回归克里格对五种气温进行插值精度验证及空间可视化。根据分析结果,得出以下结论：
（1）冬季日均气温、月均气温以及年均气温是空间非平稳的,夏季日均气温、月均气温是空间平稳的。冬季日均气温与月均气温的回归参数不同,年均气温的回归参数与冬季日均气温相同,夏季日均气温与月均气温的回归参数相同。
（2）回归克里格是在剔除趋势的基础上进行克里格插值,因此对于空间非平稳数据（存在显著的趋势）,较普通克里格能大幅度提高插值精度;而对于空间平稳数据（并不存在显著的趋势）,插值精度的提高幅度并不如空间非平稳数据显著。
（3）空间非平稳数据随着时间序列的增长,插值精度提高,然而一旦数据趋于稳定,时间序列的长短对于插值精度的影响便不再显著;空间平稳数据随着时间序列的增长,插值精度并没有显著提高。
（4）随着时间序列的增长,三种插值方法的插值精度及气温插值结果空间分布差异减小。因此,对于短时间序列的平均气温,插值方法的选择对于插值结果具有重要影响;而对于长时间序列的平均气温,不同插值方法对于插值结果的影响并不如短时间序列显著。长时间序列值是短时间序值的均值,其结果弱化了极值出现的概率,因此,冬季月均气温的低气温带分布范围与夏季月均气温的高气温带分布范围均不同程度地较对应季节的日均气温范围减小。
The authors have declared that no competing interests exist.
| [1] |
湖南复杂地形下日平均气温空间插值方法探讨 .https://doi.org/10.3969/j.issn.1000-6362.2011.04.021 URL [本文引用: 1] 摘要
基于湖南省97个常规地面气象 台站1961-2008年逐日的日平均气温资料,对湖南复杂地形下日平均气温空间插值方法进行探讨。采用考虑海拔影响下的反距离法等6种空间插值方法进行 500m×500m格距的小网格推算,对多种插值结果的误差进行了交叉检验,同时利用2007-2008年湖南省1215个自动气象站的观测资料进行误差 检验,并对不同时间段、不同区域、不同地形下的各插值方法进行了对比分析。结果表明:各空间插值方法的计算结果基本接近,但是要避免使用易导致"牛眼"现 象的IDWE等方法;在海拔高度500m以下时,除了SFE各方法误差接近,500m以上时,以IDWGE方法最优;在湘西山区以KRGE方法的空间插值 较优,湘东山区以IDWGE方法误差最小。在冬半年以MRG方法误差小,夏半年以KRGE方法误差小。相对于传统的采用界限温度的初终日进行直接插值,对 逐日平均气温插值后再推算界限温度的初终日,物理意义明确,计算结果更接近于实际值。研究结果可为湖南精细化农业气候区划提供高时空分辨率的网格资料。
Study on the spatial interpolation method for daily mean air temperature over complex terrain in Hunnan province. https://doi.org/10.3969/j.issn.1000-6362.2011.04.021 URL [本文引用: 1] 摘要
基于湖南省97个常规地面气象 台站1961-2008年逐日的日平均气温资料,对湖南复杂地形下日平均气温空间插值方法进行探讨。采用考虑海拔影响下的反距离法等6种空间插值方法进行 500m×500m格距的小网格推算,对多种插值结果的误差进行了交叉检验,同时利用2007-2008年湖南省1215个自动气象站的观测资料进行误差 检验,并对不同时间段、不同区域、不同地形下的各插值方法进行了对比分析。结果表明:各空间插值方法的计算结果基本接近,但是要避免使用易导致"牛眼"现 象的IDWE等方法;在海拔高度500m以下时,除了SFE各方法误差接近,500m以上时,以IDWGE方法最优;在湘西山区以KRGE方法的空间插值 较优,湘东山区以IDWGE方法误差最小。在冬半年以MRG方法误差小,夏半年以KRGE方法误差小。相对于传统的采用界限温度的初终日进行直接插值,对 逐日平均气温插值后再推算界限温度的初终日,物理意义明确,计算结果更接近于实际值。研究结果可为湖南精细化农业气候区划提供高时空分辨率的网格资料。
|
| [2] |
气温数据栅格化的方法及其比较 .https://doi.org/10.3321/j.issn:1007-7588.2003.06.014 URL 摘要
Direct interpolation, temperature elevation model and multiple variable regression model were respectively used to rasterize air temperature data from 592 meteorological stations in China in 1961. Air temperature data from other 58 meteorological stations were used to verify these methods. It was found that the temperature calculated by direct interpolation method had a relationship coefficient (r) of 0 95 and a standard deviation (STD) of 3 4℃ with January's mean temperature, r=0 78 and STD=3 6℃ with July's mean temperature, and r=0 87 and STD=3 3℃ with annual mean temperature respectively, that the temperature calculated by temperature elevation method had a relationship of r=0 98 and STD=2 4℃ with January's mean temperature, r=0 97 and STD=1 1℃ with July's mean temperature, and r=0 98 and STD=1 3℃ with annual mean temperature respectively, and that the temperature calculated by multiple variable regression method had a relationship of r=0 98 and STD=2 3℃ with January's mean temperature, r=0 97 and STD=1 1℃ with July's mean temperature, and r=0 98 and STD=1 4℃ with annual mean temperature respectively. Therefore, direct interpolation method is not suitable for rasterization of temperature data at large scale because of low precision, and the other two methods can be used for rasterization of temperature data.
Comparison on methods for rasterization of air temperature data. https://doi.org/10.3321/j.issn:1007-7588.2003.06.014 URL 摘要
Direct interpolation, temperature elevation model and multiple variable regression model were respectively used to rasterize air temperature data from 592 meteorological stations in China in 1961. Air temperature data from other 58 meteorological stations were used to verify these methods. It was found that the temperature calculated by direct interpolation method had a relationship coefficient (r) of 0 95 and a standard deviation (STD) of 3 4℃ with January's mean temperature, r=0 78 and STD=3 6℃ with July's mean temperature, and r=0 87 and STD=3 3℃ with annual mean temperature respectively, that the temperature calculated by temperature elevation method had a relationship of r=0 98 and STD=2 4℃ with January's mean temperature, r=0 97 and STD=1 1℃ with July's mean temperature, and r=0 98 and STD=1 3℃ with annual mean temperature respectively, and that the temperature calculated by multiple variable regression method had a relationship of r=0 98 and STD=2 3℃ with January's mean temperature, r=0 97 and STD=1 1℃ with July's mean temperature, and r=0 98 and STD=1 4℃ with annual mean temperature respectively. Therefore, direct interpolation method is not suitable for rasterization of temperature data at large scale because of low precision, and the other two methods can be used for rasterization of temperature data.
|
| [3] |
东北地区逐日气象要素的空间插值方法应用研究 .https://doi.org/10.3969/j.issn.1001-7313.2003.05.011 摘要
针对作物生长动力模型区域应用时对高精度格点逐日气象要素输入值的需要，以东北地区为例，选用克立格法(Kriging)、以经纬度分布方向为权重的距离权重反比法(IDW)及带高度梯度订正的距离权重反比法(GIDW)3种插值方法进行有限气象站点4～10月逐日气象要素空间插值方法研究，并进行插值的统计量分析和估值的交叉验证。结果表明，对温度而言，GIDW方法估值精度较高，插值结果的平滑程度适中，插值结果分布趋势也较为接近实际站点的分布。对降水而言，IDW估值精度高于Kriging，而且插值结果的平滑程度较小，更适合于日降水量的空间插值。GIDW、IDW估值精度较高的原因是研究中考虑到日最高温度、最低温度和降水量的经向、纬向梯度、海拔高度梯度存在明显的季节性变化，采用了根据气象要素经纬度方向确定权重，以及根据气象要素高度梯度年内变化进行高度订正的结果。
Spatial interpolation methods of daily weater data in northeast China. https://doi.org/10.3969/j.issn.1001-7313.2003.05.011 摘要
针对作物生长动力模型区域应用时对高精度格点逐日气象要素输入值的需要，以东北地区为例，选用克立格法(Kriging)、以经纬度分布方向为权重的距离权重反比法(IDW)及带高度梯度订正的距离权重反比法(GIDW)3种插值方法进行有限气象站点4～10月逐日气象要素空间插值方法研究，并进行插值的统计量分析和估值的交叉验证。结果表明，对温度而言，GIDW方法估值精度较高，插值结果的平滑程度适中，插值结果分布趋势也较为接近实际站点的分布。对降水而言，IDW估值精度高于Kriging，而且插值结果的平滑程度较小，更适合于日降水量的空间插值。GIDW、IDW估值精度较高的原因是研究中考虑到日最高温度、最低温度和降水量的经向、纬向梯度、海拔高度梯度存在明显的季节性变化，采用了根据气象要素经纬度方向确定权重，以及根据气象要素高度梯度年内变化进行高度订正的结果。
|
| [4] |
Comparative analysis of interpolation methods in the middle Ebro Valley (Spain): Application to annual precipitation and temperature. https://doi.org/10.3354/cr024161 URL 摘要
This paper analyzes the validity of various precipitation and temperature maps obtained by means of diverse interpolation methods. The study was carried out in an area where geographic differences and spatial climatic diversity are significant (the middle Ebro Valley in the northeast Of Spain). Two variables, annual precipitation and temperature, and several interpolation methods were used in the climate mapping: global interpolators (trend surfaces and regression models), local interpolators (Thiessen polygons, inverse distance weighting, splines), geostatistical methods (simple kriging, ordinary kriging, block kriging, directional kriging, universal kriging and co-kriging) and mixed methods (combined global, local and geostatistical methods). The validity of the maps was checked through independent test weather stations (30% of the original stations). Different statistical accuracy measurements determined the quality of the models. The results show that some interpolation methods are very similar. Nevertheless, in the case of precipitation maps, we obtained the best results using geostatistical methods and a regression model formed by 4 geographic and topographic variables. The best results for temperature mapping were obtained using the regression-based method. The accuracy measurements obtained by the different interpolation methods change significantly depending on the climatic variable mapped. The validity of interpolation methods in the creation of climatic maps, useful for agricultural and hydrologic management, is discussed.
|
| [5] |
Comparison of interpolation methods for mapping climatic and bioclimatic variables at regional scale. https://doi.org/10.1002/joc.1495 URL [本文引用: 1] 摘要
Abstract Climatic data and bioclimatic indexes have been used to study plants, animals and ecosystem distribution. GIS-based maps of climatic and bioclimatic data can be obtained by interpolating values observed at measurement stations. Since no single method can be considered as optimal for all observed regions, a major task is to propose comparisons between results obtained using different methods applied to the same data set of climate variables. We compared three methods that have been proved to be useful at regional scale: 1 - a local interpolation method based on de-trended inverse distance weighting (D-IDW), 2 - universal kriging (i.e. simple kriging with trend function defined on the basis of a set of covariates) which is optimal (i.e. BLUP, best linear unbiased predictor) if spatial association is present, 3 - multilayer neural networks trained with backpropagation (representing a complex nonlinear fitting). Long-term (1955–1990) average monthly data were obtained from weather stations measuring precipitation (201 sites) and temperature (102 sites). We analysed twelve climatic variables (temperature and precipitation) and nine bioclimatic indexes. Terrain variables and geographical location have been used as predictors of the climate variables: longitude, latitude, elevation, aspect, slope, continentality and estimated solar radiation. Based on the root mean square errors from cross-validation tests, we ranked the best method for each variable data set. Universal kriging with external drift obtained the best performances for seventeen variables of the twenty-one analysed, neural network interpolator has proven to be more efficient for three variables and D-IDW for only one. Based on these results, we used the universal kriging estimates to produce the climatic and bioclimatic maps aimed at defining the bioclimatic envelope of species. Copyright 漏 2007 Royal Meteorological Society
|
| [6] |
近10年黑龙江省气温的时空变异分析 .https://doi.org/10.3969/j.issn.1560-8999.2009.05.006 URL [本文引用: 2] 摘要
以ArcGIS Analyst为支撑,80个气象站点观测的1997-2006年的旬平均气温为插值变量,利用高程、坡向等影响气温空间分异的局地因素作为协同变量,采 用协同克里格(CoKriging)方法,考虑旬平均气温的自相关性以及旬平均气温与高程、坡向空间上的关联性,通过数据的检查、误差拟合、精度评价和模 型比较,对黑龙江省旬平均气温进行空间插值,求得全省1km×1km的各旬平均气温表面数据。36旬气温插值结果的均误差、均方根误差、平均标准差、标准 化均误差和均方根标准差的平均数分别为0.002 4℃、0.774℃、0.682℃、0.0006和1.124。由旬平均气温插值结果叠加计算出月、年平均气温表面数据。利用插值计算结果和气象站点观测 的数据,分析旬、月和年平均气温的时空分异特征,得出空间上东南部地区分异较小,其他地区分异较大;时间上11-13、12-14、19-21等旬期平均 气温有平稳下降趋势,15-17、26-28和27-29等旬期平均气温有平稳升高趋势。7月气温有稍许下降趋势,9月和11月的平均气温稍有上升趋 势,5-9月平均气温升高约1℃。年平均气温以2.9℃为均值在2.5~3.3℃之间波动,略有升高但无明显上升趋势。春季之交一些旬期平均气温变化率降 低趋稳,夏秋之交一些旬期平均气温变化率升高,实际物候有向后延迟的迹象。研究结果为气温变化监测、农业区划、土地生产潜力计算和千亿斤粮食背景下作物估 产等相关研究奠定基础。
The spatialtemporal variation analysis of air temperature in Heilongjiang province during 1997-2006 .https://doi.org/10.3969/j.issn.1560-8999.2009.05.006 URL [本文引用: 2] 摘要
以ArcGIS Analyst为支撑,80个气象站点观测的1997-2006年的旬平均气温为插值变量,利用高程、坡向等影响气温空间分异的局地因素作为协同变量,采 用协同克里格(CoKriging)方法,考虑旬平均气温的自相关性以及旬平均气温与高程、坡向空间上的关联性,通过数据的检查、误差拟合、精度评价和模 型比较,对黑龙江省旬平均气温进行空间插值,求得全省1km×1km的各旬平均气温表面数据。36旬气温插值结果的均误差、均方根误差、平均标准差、标准 化均误差和均方根标准差的平均数分别为0.002 4℃、0.774℃、0.682℃、0.0006和1.124。由旬平均气温插值结果叠加计算出月、年平均气温表面数据。利用插值计算结果和气象站点观测 的数据,分析旬、月和年平均气温的时空分异特征,得出空间上东南部地区分异较小,其他地区分异较大;时间上11-13、12-14、19-21等旬期平均 气温有平稳下降趋势,15-17、26-28和27-29等旬期平均气温有平稳升高趋势。7月气温有稍许下降趋势,9月和11月的平均气温稍有上升趋 势,5-9月平均气温升高约1℃。年平均气温以2.9℃为均值在2.5~3.3℃之间波动,略有升高但无明显上升趋势。春季之交一些旬期平均气温变化率降 低趋稳,夏秋之交一些旬期平均气温变化率升高,实际物候有向后延迟的迹象。研究结果为气温变化监测、农业区划、土地生产潜力计算和千亿斤粮食背景下作物估 产等相关研究奠定基础。
|
| [7] |
The generation of monthly gridded datasets for a range of climatic variables over the UK. https://doi.org/10.1002/joc.1161 URL [本文引用: 1] 摘要
Abstract Monthly or annual 5 km × 5 km gridded datasets covering the UK are generated for the 1961–2000 period, for 36 climatic parameters. As well as the usual elements of temperature, rainfall, sunshine, cloud, wind speed, and pressure, derived temperature variables (such as growing-season length, heating degree days, and heat and cold wave durations) and further precipitation variables (such as rainfall intensity, maximum consecutive dry days, and days of snow, hail and thunder) are analysed. The analysis process uses geographical information system capabilities to combine multiple regression with inverse-distance-weighted interpolation. Geographic and topographic factors, such as easting and northing, terrain height and shape, and urban and coastal effects, are incorporated either through normalization with regard to the 1961–90 average climate, or as independent variables in the regression. Local variations are then incorporated through the spatial interpolation of regression residuals. For each of the climatic parameters, the choice of model is based on verification statistics produced by excluding a random set of stations from the analysis for a selection of months, and comparing the observed values with the estimated values at each point. This gives some insight into the significance, direction, and seasonality of factors affecting different climate elements. It also gives a measure of the accuracy of the method at predicting values between station locations. The datasets are being used for the verification of climate modelling scenarios and are available via the Internet. 08 Crown Copyright 2005. Reproduced with the permission of Her Majesty's Stationery Office. Published by John Wiley & Sons, Ltd.
|
| [8] |
融合多源信息的地表气温插值研究 .https://doi.org/10.1007/s00376-010-0016-1 URL [本文引用: 1] 摘要
The paper presents an interpolation method for surface air temperature(SAT) based on data fusion of multiple sources.It should check whether there is a significant correlation between primary and secondary variables firsdy.Three multivariate geostatistical algorithums which includes collocated cokriging(CCK),simple kriging with varying local means (SKlm) and kriging with an external drift(KED) were introduced to incorporating ancillary information into the spatial prediction of SAT.The method was illustrated using monthly mean temperature data from more than 720 meteorological stations in China in August 2008,and cross validation was performed to evaluate the performance of the map prediction quality.The results show that:Accounting for both land surface temperature(LST) from remote sensing and digital elevation model (DEM),used as ancillary spatial information in three algorithms,outperforms accounting for only one ancillary data.Among all different methods,SKlm and KED incorporating LST and DEM have produced the best results,this is because:(1) LST is better to indicator the local trend of SAT.(2) DEM prefers to indicator the global trend of SAT.(3) Both SKlm and KED considering SAT with a non-tationary spatial distribution have better performance than others.
Surface air temperature interpolation based on multiple sources information fusion. https://doi.org/10.1007/s00376-010-0016-1 URL [本文引用: 1] 摘要
The paper presents an interpolation method for surface air temperature(SAT) based on data fusion of multiple sources.It should check whether there is a significant correlation between primary and secondary variables firsdy.Three multivariate geostatistical algorithums which includes collocated cokriging(CCK),simple kriging with varying local means (SKlm) and kriging with an external drift(KED) were introduced to incorporating ancillary information into the spatial prediction of SAT.The method was illustrated using monthly mean temperature data from more than 720 meteorological stations in China in August 2008,and cross validation was performed to evaluate the performance of the map prediction quality.The results show that:Accounting for both land surface temperature(LST) from remote sensing and digital elevation model (DEM),used as ancillary spatial information in three algorithms,outperforms accounting for only one ancillary data.Among all different methods,SKlm and KED incorporating LST and DEM have produced the best results,this is because:(1) LST is better to indicator the local trend of SAT.(2) DEM prefers to indicator the global trend of SAT.(3) Both SKlm and KED considering SAT with a non-tationary spatial distribution have better performance than others.
|
| [9] |
气象要素空间插值方法优化 .https://doi.org/10.3321/j.issn:1000-0585.2004.03.009 摘要
在区域水土平衡模型的研究中,空间插值可提供每个计算栅格的气象要素资料.本文运用反距离加权法(IDW)和梯度距离反比法(GIDW),对 1961～2000年甘肃省及其周围85个气象站点的多年平均温度与降雨量进行了内插.交叉验证结果表明:对于IDW和GIDW,二者温度插值的平均绝对误差(MAE)分别为2.28℃和0.73℃,平均相对误差(MRE)分别为29.02%和9.41%,降雨插值的MAE值依次为5.52mm和 4.90mm,MRE值分别为19.43%和17.80%,GIDW明显优于IDW.需要指出的是:对于降雨,当其经纬度和海拔高程的复相关系数大于 0.80时,GIDW插值结果优于IDW;否则相反.
Optimization of the spatial interpoation methods for climate resources. https://doi.org/10.3321/j.issn:1000-0585.2004.03.009 摘要
在区域水土平衡模型的研究中,空间插值可提供每个计算栅格的气象要素资料.本文运用反距离加权法(IDW)和梯度距离反比法(GIDW),对 1961～2000年甘肃省及其周围85个气象站点的多年平均温度与降雨量进行了内插.交叉验证结果表明:对于IDW和GIDW,二者温度插值的平均绝对误差(MAE)分别为2.28℃和0.73℃,平均相对误差(MRE)分别为29.02%和9.41%,降雨插值的MAE值依次为5.52mm和 4.90mm,MRE值分别为19.43%和17.80%,GIDW明显优于IDW.需要指出的是:对于降雨,当其经纬度和海拔高程的复相关系数大于 0.80时,GIDW插值结果优于IDW;否则相反.
|
| [10] |
Mapping the mean annual precipitation of China using local interpolation techniques. https://doi.org/10.1007/s00704-014-1105-3 URL [本文引用: 1] 摘要
Spatially explicit precipitation data are required in the research of hydrology, agriculture, ecology, and environmental sciences. In this study, two established techniques of local ordinary linear regression (OLR) and geographically weighted regression (GWR) and two new local hybrid interpolation techniques of local regression-kriging (LRK) and geographically weighted regression kriging (GWRK) were compared to predict the spatial distribution of mean annual precipitation of China. Precipitation data from 684 meteorological stations were used in the analysis, and a stepwise regression analysis was used to select six covariates, including longitude, latitude, elevation, slope, surface roughness, and river density. The four spatial prediction methods (OLR, GWR, LRK, and GWRK) were implemented with local regression techniques with different number of neighbors (50, 100, 150, and 200). The prediction accuracy was assessed at validation sites with the root mean squared deviation, mean estimation error, and R -square values. The results showed that LRK outperforms OLR and GWRK outperforms GWR, indicating that adding the kriging of regression residuals can help improve the prediction performance. GWRK gives the best prediction but the accuracy of estimation varies with the number of neighborhood points used for modeling. Although LRK is outperformed by GWRK, LRK is still recommended as a powerful and practical interpolation method given its computation efficiency. However, if LRK and GWRK are used to extrapolate prediction values, post-processing in the areal interpolation will be needed.
|
| [11] |
气温插值中不同空间插值方法的适用性分析: 以江苏省为例 .https://doi.org/10.3724/SP.J.1047.2011.00539 URL [本文引用: 1] 摘要
Air temperature is an important parameter observed in metrological stations,and there are many ways to improve the precision of air temperature interpolation result.In this paper by using the air temperature data at 67 meteorological stations in Jiangsu Province in the year of 2003 and digital elevation model(DEM) data with spatial resolution of 30mx30m,four common interpolation methods,including Inverse Distance Weighting(IDW),Spline with tension(Spline),Ordinary Kriging(OK) and Co-Kriging(CK),were used to interpolate the monthly and yearly mean air temperature and the precision of those four methods was compared by using cross validation method.The results showed that OK has a much higher precision than IDW and Spline,indicating the method accounting for spatial self-correlation is more accurate than others.Four auxiliary variables,including latitude,longitude,distance from the coast and elevation,were selected for CK,and correlation analysis showed that the monthly mean air temperatures are best correlated with latitude,and the three other variables followed.As the four variables are correlated with each other,principal component analysis(PCA) was conducted in this paper.The first principal component mainly representing longitude and distance from the coast and the second one mainly representing latitude were utilized as the optimized auxiliary variables for Co-Kriging interpolation in most months except March whose input is only the second one,April and July whose inputs are the second and the fourth principal components.The results indicated that the precision of CK which makes good use of related auxiliary factors is slightly higher than that of OK;while it is obviously better than OK where there are fewer stations and is a potential ideal method for air temperature interpolation.The results of this paper also showed that distance from the coast is a critical factor to the spatial pattern of air temperature in Jiangsu,China,which should be an auxiliary variable for CK.
Suitability assessment of different interpolation methods in the gridding process of station collected air temperature: A case study in Jiangsu province, China .https://doi.org/10.3724/SP.J.1047.2011.00539 URL [本文引用: 1] 摘要
Air temperature is an important parameter observed in metrological stations,and there are many ways to improve the precision of air temperature interpolation result.In this paper by using the air temperature data at 67 meteorological stations in Jiangsu Province in the year of 2003 and digital elevation model(DEM) data with spatial resolution of 30mx30m,four common interpolation methods,including Inverse Distance Weighting(IDW),Spline with tension(Spline),Ordinary Kriging(OK) and Co-Kriging(CK),were used to interpolate the monthly and yearly mean air temperature and the precision of those four methods was compared by using cross validation method.The results showed that OK has a much higher precision than IDW and Spline,indicating the method accounting for spatial self-correlation is more accurate than others.Four auxiliary variables,including latitude,longitude,distance from the coast and elevation,were selected for CK,and correlation analysis showed that the monthly mean air temperatures are best correlated with latitude,and the three other variables followed.As the four variables are correlated with each other,principal component analysis(PCA) was conducted in this paper.The first principal component mainly representing longitude and distance from the coast and the second one mainly representing latitude were utilized as the optimized auxiliary variables for Co-Kriging interpolation in most months except March whose input is only the second one,April and July whose inputs are the second and the fourth principal components.The results indicated that the precision of CK which makes good use of related auxiliary factors is slightly higher than that of OK;while it is obviously better than OK where there are fewer stations and is a potential ideal method for air temperature interpolation.The results of this paper also showed that distance from the coast is a critical factor to the spatial pattern of air temperature in Jiangsu,China,which should be an auxiliary variable for CK.
|
| [12] |
基于DEM的中国陆地多年平均温度插值方法 .https://doi.org/10.3321/j.issn:0375-5444.2004.03.006 URL [本文引用: 1] 摘要
Statistical interpolation of the temperature for the missing points is one of the most popular approaches for generating high spatial resolution data sets. However, many interpolation methods used by previous studies are purely mathematic ways, without geographical significance being considered. In the present study the authors interpolate the monthly and annual mean temperature climatologies using 726-station observations in China, utilizing improved methods by taking into account geographical factors such as latitude, longitude, altitude. In addition, a smart distance-searching technique is adopted, which helps select the optimum stations on which the guess values at missing points are generated. Results show that the methods used here have evident advantages over the previous approaches. The mean absolute err of ordinary inverse-distance-squared (IDS) technique is in the range of 1.44-1.63oC, on average 1.51oC. The smart distance searching technique yield a MAE of 0.53-0.92oC, on average 0.69oC. Errors have been reduced as much as 50%.
Smart distance searching-based and DEM-informed interpolation of surface air temperature in China. https://doi.org/10.3321/j.issn:0375-5444.2004.03.006 URL [本文引用: 1] 摘要
Statistical interpolation of the temperature for the missing points is one of the most popular approaches for generating high spatial resolution data sets. However, many interpolation methods used by previous studies are purely mathematic ways, without geographical significance being considered. In the present study the authors interpolate the monthly and annual mean temperature climatologies using 726-station observations in China, utilizing improved methods by taking into account geographical factors such as latitude, longitude, altitude. In addition, a smart distance-searching technique is adopted, which helps select the optimum stations on which the guess values at missing points are generated. Results show that the methods used here have evident advantages over the previous approaches. The mean absolute err of ordinary inverse-distance-squared (IDS) technique is in the range of 1.44-1.63oC, on average 1.51oC. The smart distance searching technique yield a MAE of 0.53-0.92oC, on average 0.69oC. Errors have been reduced as much as 50%.
|
| [13] |
自然地理要素空间插值的几个问题 .https://doi.org/10.3321/j.issn:1000-0585.2004.04.001 URL [本文引用: 1] 摘要
The spatial interpolation of some physical geographical elements is becoming increasingly important nowadays in resources management, disaster control, environment improvement and the research of global change. The core of the spatial interpolation of those elements is to seek the functions that can reveal their characteristics of spatial distribution. But, as for specified regions and sample data, there are many functions in the list for choice. And the best choice is difficult to make because of the complex effects from theoretical foundation, algorithm, temporal spatial scale, and attributes of sample data. By referring to the major achievements in the interpolation research field, this paper comes to the point that the accuracy of certain spatial interpolation depends on its capability of reflecting the element's spatial variance and spatial correlation. Models using other elements as variables, when regression variable has high correlation with interpolation variable, will give more accurate results than others, because they have better reflection of spatial variance. Models without other element variables change in accuracy according to their consideration of the anisotropic characteristics or not. With spatial temporal scales' variance, the disposed spatial variance and correlations will be different, which affects the interpolation accuracy. The density, spatial distribution, data extent of sample points also makes the interpolation results different for the same reason. As for applications, the optimal interpolation method should be picked out after the analysis of those spatial characteristics embedded in the sample dataset.
Problems of the spatial interpolation of physical geographical elements. https://doi.org/10.3321/j.issn:1000-0585.2004.04.001 URL [本文引用: 1] 摘要
The spatial interpolation of some physical geographical elements is becoming increasingly important nowadays in resources management, disaster control, environment improvement and the research of global change. The core of the spatial interpolation of those elements is to seek the functions that can reveal their characteristics of spatial distribution. But, as for specified regions and sample data, there are many functions in the list for choice. And the best choice is difficult to make because of the complex effects from theoretical foundation, algorithm, temporal spatial scale, and attributes of sample data. By referring to the major achievements in the interpolation research field, this paper comes to the point that the accuracy of certain spatial interpolation depends on its capability of reflecting the element's spatial variance and spatial correlation. Models using other elements as variables, when regression variable has high correlation with interpolation variable, will give more accurate results than others, because they have better reflection of spatial variance. Models without other element variables change in accuracy according to their consideration of the anisotropic characteristics or not. With spatial temporal scales' variance, the disposed spatial variance and correlations will be different, which affects the interpolation accuracy. The density, spatial distribution, data extent of sample points also makes the interpolation results different for the same reason. As for applications, the optimal interpolation method should be picked out after the analysis of those spatial characteristics embedded in the sample dataset.
|
| [14] |
气象要素空间插值方法的比较分析 .
Climate factor is not only considered as important index bases in the comprehensive and sequential classes,but also has great impact on composing and growth of species,accumulation and succession of dry-substance,and also has close relation with biological diversity and potential of land utilization.The character of climatic spatial distribution indicates that the result of average annual air temperature,average annual precipitation and annual accumulated temperature becomes to decrease with the increasing of latitude and elevation.But there is no clear difference between spatial interpolations which is calculated by different interpolation method.Based on ArcMap8.3 GIS,the observed data which are average annual air temperature,average annual accumulated temperature and average annual precipitation for 30 years from 1961 to 1990 from 2114 meteorological stations in China and its nearby regions are compared and analyzed by Spline,IDS,OK.The result of cross-validation tests shows that the precision of interpolated result is very high when select proper meteorological stations.In ordinary kriging,the precision of interpolated result is similar among Spherical Model,Circular Model,Exponential Model,except Gaussian Model with the same climatic factor.The relative mean errors of these three methods are ranked as: OK Spline IDS,relative average errors are 7.65%,7.9% and 7.95% for interpolating average annual air temperature.Relative excellence is OK Spline = IDS,relative average errors are 8.31%,8.76% and 8.76% for average annual precipitation.OK IDS Spline for average annual accumulated temperature,and relative excellence is 5.82%,6.11% and 6.13%.
Analyze and compare the spatial interpolation methods for climate factor.
Climate factor is not only considered as important index bases in the comprehensive and sequential classes,but also has great impact on composing and growth of species,accumulation and succession of dry-substance,and also has close relation with biological diversity and potential of land utilization.The character of climatic spatial distribution indicates that the result of average annual air temperature,average annual precipitation and annual accumulated temperature becomes to decrease with the increasing of latitude and elevation.But there is no clear difference between spatial interpolations which is calculated by different interpolation method.Based on ArcMap8.3 GIS,the observed data which are average annual air temperature,average annual accumulated temperature and average annual precipitation for 30 years from 1961 to 1990 from 2114 meteorological stations in China and its nearby regions are compared and analyzed by Spline,IDS,OK.The result of cross-validation tests shows that the precision of interpolated result is very high when select proper meteorological stations.In ordinary kriging,the precision of interpolated result is similar among Spherical Model,Circular Model,Exponential Model,except Gaussian Model with the same climatic factor.The relative mean errors of these three methods are ranked as: OK Spline IDS,relative average errors are 7.65%,7.9% and 7.95% for interpolating average annual air temperature.Relative excellence is OK Spline = IDS,relative average errors are 8.31%,8.76% and 8.76% for average annual precipitation.OK IDS Spline for average annual accumulated temperature,and relative excellence is 5.82%,6.11% and 6.13%.
|
| [15] |
逐日气象要素空间插值方法的比较 .URL 摘要
采用距离反比权重法(IDW)、协克里格法(CK)和薄盘样条法 (TPS)3种不同的空间插值方法,对我国1951-2005年气象数据完整的559个气象站点逐月第15日的平均基本气象要素(最高气温、最低气温、日照时数和降水量)进行了插值分析与评价.结果表明:3种插值方法中,TPS法对最高气温和最低气温插值的根均方差(RMSE)最小(1.02 ℃和1.12 ℃)、R~2最大(0.9916和0.9913);不同季节中,TPS法对秋季最高气温、夏季最低气温进行插值的RMSE均最小(0.83℃、 0.86℃),R~2均为秋季最高.对于日照时数和降水量而言,TPS法的RMSE最小(0.59 h和1.01 mm)、R~2最大(0.9118和0.8135);不同季节中,TPS法对冬季日照时数进行插值的RMSE最小(0.49 h)、R~2最大(0.92
Comparison of spatial interpolation methods for daily meteorological elements. URL 摘要
采用距离反比权重法(IDW)、协克里格法(CK)和薄盘样条法 (TPS)3种不同的空间插值方法,对我国1951-2005年气象数据完整的559个气象站点逐月第15日的平均基本气象要素(最高气温、最低气温、日照时数和降水量)进行了插值分析与评价.结果表明:3种插值方法中,TPS法对最高气温和最低气温插值的根均方差(RMSE)最小(1.02 ℃和1.12 ℃)、R~2最大(0.9916和0.9913);不同季节中,TPS法对秋季最高气温、夏季最低气温进行插值的RMSE均最小(0.83℃、 0.86℃),R~2均为秋季最高.对于日照时数和降水量而言,TPS法的RMSE最小(0.59 h和1.01 mm)、R~2最大(0.9118和0.8135);不同季节中,TPS法对冬季日照时数进行插值的RMSE最小(0.49 h)、R~2最大(0.92
|
| [16] |
基于GIS的新疆气温数据栅格化方法研究 .URL 摘要
With Surfer,Grads as a platform for direct space interpolation was widely used in meteorological rasterization of air temperature data,whatever the spatial interpolation technique(Spline,IDW,Lagrangian,Hennite interpolation,etc.),do not take into account the effects of topography on the air temperature distribution,In recent years with the expansion of GIS technology applications,the method of regression model by geographic factors(elevation,longitude,latitude,etc.) combined with spatial interpolation was used in grid-based regional climate factors and get good results.In this paper,used regression analysis methods combined with GIS spatial interpolation to rasterization of year mean air temperatures from 1971 to 2010 in Xinjiang area,the 99 meteorological stations(10 of them in order to verify) that has complete observations involved in the calculation.We use the following method for air temperature data rasterization in Xinjiang region,Firstly,establish the average temperature multiple regression model with the air temperature data that measured by weather station(excluding test station) for the output variables,and the longitude grid data,latitude grid data and altitude grid data of meteorological stations for the input variables,obtain the regression equation and the temperature residuals data for each weather station;Secondly,calculate the air temperature grid data(regular part) of the each observations station use the digital elevation model(DEM),the latitude grid data and longitude grid data by the regression equation,and then the residuals grid data(irregular part) to be rasterized with a spatial interpolation method(Three methods including IDW,Kriging and Spline);Finally,the two parts of the data grid computing has been to estimate the actual temperature.The authors used this method to rasterization the air temperature grid data of the Xinjiang region for many years(the average temperature data for 40 years,most recently 2010,the hottest years 2007 and the coldest years 1984).Comparative and analysis of residual data interpolation with inverse distance weighting method(IDW),ordinary kriging(Kriging) and Spline(Spline) method,overall,the result of mean absolute errors(MAE) and Root Mean Squared Interpolation Error(RMSIE) from cross-validation tests is IDW gives lowest errors.The other question is,even if we use the best method(IDW) to create raster data of the air temperature in Xinjiang,the rasterized grid of error is significantly larger than the most other provinces in China,mainly due to Xinjiang's unique geospatial and sparse distribution of meteorological observation site,so how to improve the accuracy of simulation grid is the key of the future rasterized grid in Xinjiang.
GIS-based spatial interpolation of air temperature in Xinjiang. URL 摘要
With Surfer,Grads as a platform for direct space interpolation was widely used in meteorological rasterization of air temperature data,whatever the spatial interpolation technique(Spline,IDW,Lagrangian,Hennite interpolation,etc.),do not take into account the effects of topography on the air temperature distribution,In recent years with the expansion of GIS technology applications,the method of regression model by geographic factors(elevation,longitude,latitude,etc.) combined with spatial interpolation was used in grid-based regional climate factors and get good results.In this paper,used regression analysis methods combined with GIS spatial interpolation to rasterization of year mean air temperatures from 1971 to 2010 in Xinjiang area,the 99 meteorological stations(10 of them in order to verify) that has complete observations involved in the calculation.We use the following method for air temperature data rasterization in Xinjiang region,Firstly,establish the average temperature multiple regression model with the air temperature data that measured by weather station(excluding test station) for the output variables,and the longitude grid data,latitude grid data and altitude grid data of meteorological stations for the input variables,obtain the regression equation and the temperature residuals data for each weather station;Secondly,calculate the air temperature grid data(regular part) of the each observations station use the digital elevation model(DEM),the latitude grid data and longitude grid data by the regression equation,and then the residuals grid data(irregular part) to be rasterized with a spatial interpolation method(Three methods including IDW,Kriging and Spline);Finally,the two parts of the data grid computing has been to estimate the actual temperature.The authors used this method to rasterization the air temperature grid data of the Xinjiang region for many years(the average temperature data for 40 years,most recently 2010,the hottest years 2007 and the coldest years 1984).Comparative and analysis of residual data interpolation with inverse distance weighting method(IDW),ordinary kriging(Kriging) and Spline(Spline) method,overall,the result of mean absolute errors(MAE) and Root Mean Squared Interpolation Error(RMSIE) from cross-validation tests is IDW gives lowest errors.The other question is,even if we use the best method(IDW) to create raster data of the air temperature in Xinjiang,the rasterized grid of error is significantly larger than the most other provinces in China,mainly due to Xinjiang's unique geospatial and sparse distribution of meteorological observation site,so how to improve the accuracy of simulation grid is the key of the future rasterized grid in Xinjiang.
|
| [17] |
基于GIS技术的青海省最低气温空间插值方法探讨 .https://doi.org/10.1080/00949651003724790 URL [本文引用: 1] 摘要
The multiple linear regression model and multiple linear regression model were established using Kriging interpolation method and Statistic Analysis Software based on the climate data in 30 years from 50 meteorological stations in Qinghai Province.Grid cell value with 500 m x 500 m of minimum temperature in January,April,July,October were calculated to produce a minimum temperature map of Qinghai Province.The effects of different interpolation methods were analyzed.Results showed that the multiple linear regression model and multi-dimensional quadratic trend surface model were better than the Kriging interpolation method in January,April and October;and Kriging interpolation method was better in July.
GIS technology based minimum temperature spatial interpolation method in Qinghai province. https://doi.org/10.1080/00949651003724790 URL [本文引用: 1] 摘要
The multiple linear regression model and multiple linear regression model were established using Kriging interpolation method and Statistic Analysis Software based on the climate data in 30 years from 50 meteorological stations in Qinghai Province.Grid cell value with 500 m x 500 m of minimum temperature in January,April,July,October were calculated to produce a minimum temperature map of Qinghai Province.The effects of different interpolation methods were analyzed.Results showed that the multiple linear regression model and multi-dimensional quadratic trend surface model were better than the Kriging interpolation method in January,April and October;and Kriging interpolation method was better in July.
|
| [18] |
中国东北地区气象数据的空间平稳性检验 .Spatial stationarity tests on meteorological data of north-east area of China. |
/
| 〈 |
|
〉 |
