气候变暖及人类活动加剧所带来的干旱问题，长期以来都是气候研究的重点之一。目前卫星遥感技术在监测全球区域干旱方面发挥着重要作用，但一直受限于云层覆盖的干扰、卫星重访周期的限制，以及传感器设计的局限。为了提升干旱监测的效能和准确性。本文通过时空数据融合技术，获取高时空分辨率数据集。以更准确的揭示区域干旱的时空分布特征及其演变规律，为深入理解和应对干旱问题提供有力支持。

本研究以富平县为研究区，选取2013至2022年的Landsat-8和MODIS遥感影像。为了进一步提升数据的时空分辨率，运用了STARFM（Space-Time Adaptive Reflectance Fusion Model）、FSDAF（Flexible Spatiotemporal Data Fusion）、FSDAF2、OL-FSDAF（Object Level-FSDAF）、OL-FSDAF2五种模型，得到了8天、30米的NDVI与LST影像，进一步反演得到了TVDI数据；并划分干旱等级，全面剖析了富平县近十年的干旱时空特征；最后探讨了气象因素、地形、人类活动等方面对富平县干旱的影响。研究结果表明：

（1）以Landsat-9卫星获取对应时间的NDVI和LST数据作为参考标准，使用6种评价指标（AAD、PCCs、RMSE、SAM、SSIM和ERGAS）对5种不同融合模型生成的8天周期NDVI和LST数据集进行了精度评价。结果表明，OL-FSDAF2模型在NDVI数据中PCCs（0.8565），SSIM（0.8755）和ERGAS（18.0765）展现出显著的优势，而在LST数据中，OL-FSDAF2模型仅在PCCs（0.8393）指标上表现最佳，其他指标只是相对较好。然后，利用2013至2022年间0-10cm的土壤相对湿度（RSM）数据，对这五种融合模型所生成的TVDI数据进行了线性回归分析。分析结果显示，OL-FSDAF2模型生成的TVDI呈现出较高的R²（0.7943）值，表明其具有较高的数据拟合精度。

（2）基于融合效果最佳的OL-FSDAF2方法分析了富平县近十年来的干旱时空变化特征。结果表明，富平县常受持续时间长，影响范围广的干旱影响，且以干旱等级为主。2020年旱情最为严重，影响范围占区域总面积的23.09%，为近十年之最，2014年次之。相比之下，2019年的干旱影响范围最少。在2013年至2022年这一时期，逐8天干旱出现了明显的波动。在大多数年份，夏季是区域干旱的主要驱动力。这主要是由于富平县五月份降水稀少，降雨时间晚，再加上春季增温速度快、风力强等气象因素的影响，导致土壤水分快速流失，容易发生干旱灾害；六月和八月的旱情更为严峻，夏季气温的升高伴随着降水的减少，干旱现象随之出现，对农作物生长和产量构成了直接影响。富平县67%的地域TVDI数值呈现下降趋势，旱情程度缓解；西部、北部和中部偏北的部分地区呈上升趋势，旱情加剧。70%的地区TVDI变化轻微上升或轻微下降，而中部北面和北部区域TVDI变化相对显著和强显著，显著变化地区遍布全县。

（3）基于富平县气候变化、地形以及土地利用对干旱变化的影响因素进行了分析。干旱与海拔、坡度整体呈现负相关的趋势，土地利用的改变可能导致干旱情况发生变化，整体趋势是水田、林地、草地、旱地的TVDI数值逐渐增大。在气象方面，气温和日照时数对于TVDI数值有促进作用，而降水和蒸散量与TVDI呈负相关。日照对TVDI的影响程度高于其他气象因素。

The issue of drought, exacerbated by global warming and escalating anthropogenic interventions, has been a longstanding subject of climatic inquiries. Nowadays, while satellite remote sensing technology holds immense importance in tracking droughts worldwide and in specific locales, its effectiveness is often hampered by factors like cloud obstruction, infrequent satellite visits, and inherent sensor limitations. Seeking to bolster the precision and reliability of drought tracking, this treatise utilizes advanced spatiotemporal data fusion methods to compile a comprehensive high-resolution dataset. This approach facilitates a more nuanced understanding of how drought patterns evolve across both time and space, bolstering capacity to grasp and tackle drought-related challenges.

This investigation has zeroed in on Landsat-8 and MODIS satellite imagery from Fuping spanning the years 2013 to 2022. To elevate the data's spatiotemporal granularity further, deployed five distinct models: STARFM (Space-Time Adaptive Reflectance Fusion Model), FSDAF (Flexible Spatiotemporal Data Fusion）, FSDAF2, OL-FSDAF (Object Level-FSDAF), OL-FSDAF2. These models aided in generating NDVI and LST images with an 8-day temporal frequency and 30-meter spatial clarity. Leveraging these images, derived TVDI metrics, categorized drought intensities, and conducted an exhaustive analysis of Fuping's drought patterns over the past ten years. Moreover, this study delves into how various meteorological factors, topographical features, and human endeavors influence agricultural droughts in Fuping. The research findings reveal:

(1) Using NDVI and LST data obtained from the Landsat-9 satellite at corresponding times as a reference standard, six evaluation metrics (AAD, PCCs, RMSE, SAM, SSIM, and ERGAS) were applied to assess the accuracy of 8-day NDVI and LST datasets generated by five different fusion models. The results showed that the OL-FSDAF2 model demonstrated significant advantages in terms of PCCs (0.8565), SSIM (0.8755), and ERGAS (18.0765) for NDVI data. However, for LST data, the OL-FSDAF2 model only excelled in the PCCs (0.8393) metric, with relatively good performance in other metrics. Subsequently, a linear regression analysis was conducted on TVDI data generated by the five fusion models using soil relative humidity (RSM) data from 0-10cm between 2013 and 2022. The analysis revealed that the TVDI generated by the OL-FSDAF2 model exhibited a higher R2 (0.7943) value, indicating its superior data fitting accuracy.

(2) With the utilization of the superior OL-FSDAF2 method, an examination of drought spatio-temporal patterns in Fuping County over the past ten years was undertaken. The findings revealed that Fuping County frequently experiences persistent and widespread droughts, often of moderate severity. Notably, 2020 witnessed the most extreme drought, affecting 23.09% of the county, closely followed by 2014, while 2019 saw the least drought. Every eight days, fluctuations in drought conditions were observed. Droughts tend to worsen in summer, primarily due to factors like limited precipitation in May, delayed rainfall, rapid spring warming, and strong winds, all leading to soil moisture depletion and elevated spring drought risks. Droughts are typically more pronounced in June and August, when rising temperatures and diminishing precipitation directly affect crop growth and production. By applying Theil-Sen Median trend analysis and the Mann-Kendall test to seasonal and annual average TVDI data from 2013 to 2022, it was noted that 67% of Fuping County showed a downward TVDI trend, suggesting drought alleviation. However, an upward trend was observed in the western, northern, and central-northern regions, indicating deteriorating drought situations. While TVDI shifts were subtle in 70% of the areas, they were significant in the northern and central regions, showcasing county-wide disparities.

(3) Considering climatic shifts, terrain, and land utilization in Fuping County, an analysis of the factors influencing drought variations was conducted. Agricultural drought in Fuping County demonstrates an inverse correlation with altitude and slope, and alterations in land use patterns can influence agricultural drought trends. Typically, TVDI values for paddy fields, forests, meadows, and drylands exhibit an upward trend. From a meteorological standpoint, temperature and sunshine duration positively impact the Vegetation Drought Index (TVDI), whereas precipitation and evapotranspiration correlate negatively with TVDI. Notably, sunshine has a more profound influence on soil moisture than other climatic elements.

[1] Mishra A K, Singh V P. A review of drought concepts[J]. Journal of Hydrology, 2010, 391(1-2): 202-216.

[2] 叶笃正. 中国的全球变化预研究[M] 地震出版社,1992: 5-11.

[3] Tfwala C M, Rensburg L V, Schall R, et al. Drought dynamics and interannual rainfall variability on the Ghaap plateau, South Africa, 1918-2014[J]. Physics and Chemistry of The Earth, 2018, 107(1474-7065): 1-7.

[4] Gao X, Zhao Q, Zhao X, et al. Temporal and spatial evolution of the standardized precipitation evapotranspiration index (SPEI) in the Loess Plateau under climate change from 2001 to 2050[J]. Science of the Total Environment, 2017, 595(0048-9697): 191-200.

[5] Wilhite D A, Glantz M H. Understanding: the Drought Phenomenon: The Role of Definitions: Water International[J]. Water International, 1985, 10(3): 111–120.

[6] 陈怀亮, 张红卫, 刘荣花, 等. 中国农业干旱的监测, 预警和灾损评估[J]. 科技导报, 2009, (11): 82-92.

[7] 闫娜, 李登科, 杜继稳, 等. 基于MODIS产品LST/NDVI/EVI的陕西旱情监测[J]. 自然灾害学报, 2010, 19(4): 178-182.

[8] 杨丽萍, 陈发虎, 颉耀文. 国内多源遥感影像信息融合技术的新进展[J]. 遥感技术与应用, 2007, 22(1): 116-122.

[9] 李柏贞, 周广胜. 干旱指标研究进展[J]. 生态学报, 2014, 34(5): 1043-1052.

[10] 佟斯琴, 张继权, 哈斯, 等. 基于MOD16的锡林郭勒草原14年蒸散发时空分布特征[J]. 中国草地学报, 2016, (4): 83-91.

[11] 张静, 任志远. 基于MOD16的汉江流域地表蒸散发时空特征[J]. 地理科学, 2017, 37(2): 274-282.

[12] Anyamba A, Tucker C J. Analysis of Sahelian vegetation dynamics using NOAA-AVHRR NDVI data from 1981–2003[J]. Journal of Arid Environments, 2005, 63(3): 596-614.

[13] Yin R S, Xiang Q, Xu J T, et al. Modeling the Driving Forces of the Land Use and Land Cover Changes Along the Upper Yangtze River of China[J]. Environmental Management, 2010, 45(3): 454-465.

[14] 胡猛, 冯起, 席海洋. 遥感技术监测干旱区土壤水分研究进展[J]. 土壤通报, 2013, 44 (5): 1270-1275.

[15] 汪潇, 张增祥, 赵晓丽, 等. 遥感监测土壤水分研究综述[J]. 土壤学报, 2007, 44(1): 157-163.

[16] Ghassemian, Hassan. A review of remote sensing image fusion methods[J]. Information Fusion, 2016, 32: 75-89.

[17] 牛海鹏, 王占奇, 肖东洋. 基于时空数据融合的县域水稻种植面积提取[J]. 农业机械学报, 2020, 51(4): 156-163.

[18] 张立福, 彭明媛, 孙雪剑, 等. 遥感数据融合研究进展与文献定量分析(1992-2018)[J]. 遥感学报, 2019, 23(4): 603-619.

[19] Bai L, Long D, Yan L. Estimation of Surface Soil Moisture With Downscaled Land Surface Temperatures Using a Data Fusion Approach for Heterogeneous Agricultural Land[J]. Water resources research, 2019, 55(2): 1105-1128.

[20] Dong T, Liu J, et al. Estimating winter wheat biomass by assimilating leaf area index derived from fusion of Landsat-8 and MODIS data[J]. 2016, 49: 63-74.

[21] Liao C, Wang J. Evaluation of spatio-temporal data fusion methods for generating NDVI time series in cropland areas[C]. IGARSS 2016 - 2016 IEEE International Geoscience and Remote Sensing Symposium, 2016: 2570-2573.

[22] Mityok Z K, Bolton D K, Coops N C, et al. Snow cover mapped daily at 30 meters resolution using a fusion of multi-temporal MODIS NDSI data and Landsat surface reflectance[J]. Canadian Journal of Remote Sensing, 2018, 44(5): 413-434.

[23] 王丽涛, 王世新, 周艺, 等. 旱情遥感监测研究进展与应用案例分析[J]. 遥感学报, 2011, 15(6): 1315-1330.

[24] 侯英雨, 何延波, 柳钦火, 等. 干旱监测指数研究[J]. 生态学杂志, 2007, 26(6): 892-897.

[25] Watson K. Application of thermal modeling in the geologic interpretation of IR images[J]. Proceedings of the IEEE, 1975, 63(1): 128-137.

[26] Price, John C. Thermal inertia mapping: A new view of the Earth[J]. Journal of Geophysical Research, 1977, 82(18): 2582-2590.

[27] Price J C. On the analysis of thermal infrared imagery: The limited utility of apparent thermal inertia[J]. Remote Sensing of Environment, 1985, 18(1): 59-73.

[28] 余涛, 田国良. 热惯量法在监测土壤表层水分变化中的研究[J]. 遥感学报, 1997, 1(1): 24-31.

[29] 陈怀亮, 冯定原, 邹春辉, 等. 用NOAA/AVHRR资料遥感土壤水分时风速的影响[J]. 南京气象学院学报, 1999, 22(2): 219-224.

[30] 刘良明, 李德仁. 基于辅助数据的遥感干旱分析[J]. 武汉大学学报(信息科学版), 1999, 24(4): 300-305.

[31] 李星敏, 刘安麟, 张树誉, 等. 热惯量法在干旱遥感监测中的应用研究[J]. 干旱地区农业研究, 2005, 23(1): 54-59.

[32] 杨玉永, 徐秀杰, 杨丽萍. 墒情遥感监测中热惯量模型的修正[J]. 灌溉排水学报, 2018, 37(6): 54-59.

[33] Rouse J W, Haas R H, Schell J A, et al. Monitoring Vegetation Systems in the Great Plains with Erts[J]. NASA Special Publication, 1974, 351(1): 309-329.

[34] Kogan, F. N. Remote sensing of weather impacts on vegetation in non-homogeneous areas[J]. International Journal of Remote Sensing, 1990, 11(8): 1405-1419.

[35] Liu W T, Kogan F N. Monitoring regional drought using the Vegetation Condition Index[J]. 17(14): 2761-2782.

[36] 冯强, 田国良, 王昂生. 基于植被状态指数的全国干旱遥感监测试验研究(Ⅰ)——资料分析与处理部分[J]. 干旱区地理, 2004, 27(2): 131-136.

[37] 陈维英, 肖乾广, 盛永伟. 距平植被指数在1992年特大干旱监测中的应用[C].气象卫星遥感反演和应用学术会议, 1994: 106-112.

[38] 杜晓. 植被叶面水遥感监测及其时空特征分析[D].北京: 中国科学院研究生院(遥感应用研究所), 2006.

[39] 姚远, 陈曦, 钱静. 遥感数据在农业旱情监测中的应用研究进展[J]. 光谱学与光谱分析, 2019, 39(4): 1005-1012.

[40] 高峰, 王介民, 孙成权, 等. 微波遥感土壤湿度研究进展[J]. 遥感技术与应用, 2001, 16(2): 97-102.

[41] 李震, 郭华东, 施建成. 综合主动和被动微波数据监测土壤水分变化[J]. 遥感学报, 2002, 6(6): 481-484.

[42] 王磊, 文军, 张堂堂, 等. 卫星被动微波遥感土壤湿度研究进展[J]. 气象科技, 2009, 37(1): 67-73.

[43] 周鹏, 丁建丽, 王飞, 等. 植被覆盖地表土壤水分遥感反演[J]. 遥感学报, 2010, 14(5): 959-973.

[44] 余凡, 赵英时. 合成孔径雷达反演裸露地表土壤水分的新方法[J]. 武汉大学学报：信息科学版, 2010, 35(3): 319-321.

[45] Goward S N, Hope A S. Evapotranspiration from combined reflected solar and emitted terrestrial radiation - Preliminary FIFE results from AVHRR data[J]. Advances in Space Research, 1989, 9(7): 239-249.

[46] Ramakrishna N, Pierce L, et al. Developing Satellite-derived Estimates of Surface Moisture Status[J]. Journal of Applied Meteorology, 1993, 32(3): 548-557.

[47] Sandholt I, Rasmussen K, Andersen J. A simple interpretation of the surface temperature/vegetation index space for assessment of surface moisture status[J]. Remote Sensing of Environment, 2002, 79(2–3): 213-224.

[48] Liang L, Zhao S H, Qin Z H, et al. Drought Change Trend Using MODIS TVDI and Its Relationship with Climate Factors in China from 2001 to 2010[J]. Journal of Integrative Agriculture(农业科学学报(英文)), 2014, 1(7): 1501-1508.

[49] 王海, 杨祖祥, 王麟, 等. TVDI在云南2009/2010年干旱监测中的应用[J]. 云南大学学报(自然科学版), 2014, 36(1): 59-65.

[50] 王婷婷, 张洪岩, 郭笑怡, 等. 基于温度植被干旱指数的松辽平原干旱时空特征[J]. 干旱区研究, 2014, 31(3): 383-389.

[51] 王植, 陈炟君, 董斌. 基于TVDI的2013—2017年沈阳市旱情等级评估[J]. 测绘与空间地理信息, 2018, 41(11): 12-17.

[52] 温国涛, 白建军, 孙嵩松. 基于时间序列遥感数据的陕西省2004-2014年干旱变化特征分析[J]. 干旱地区农业研究, 2018, 36(1): 221-229.

[53] 吴英杰, 全强, 陈晓俊, 等. 2000—2018年锡林郭勒地区干旱时空变化及其气候响应[J]. 2020, 43(5): 1289-1397.

[54] 康尧, 郭恩亮, 王永芳, 等. 温度植被干旱指数在蒙古高原干旱监测中的应用[J]. 应用生态学报, 2021, 32(7): 2534-2544.

[55] 覃艺, 张廷斌, 易桂花, 等. 2000年以来内蒙古生长季旱情变化遥感监测及其影响因素分析[J]. 自然资源学报, 2021, 36(2): 459-475.

[56] 黄波. 时空遥感影像融合研究的进展与趋势[J]. 四川师范大学学报：自然科学版, 2020, 43(4): 427-434.

[57] Feng G, Masek J G, Schwaller M R, et al. On the Blending of the Landsat and MODIS Surface Reflectance: Predicting Daily Landsat Surface Reflectance[J]. IEEE Transactions on Geoscience and Remote Sensing, 2006, 44(8): 2207-2218.

[58] Weng Q, Fu P, Gao F. Generating daily land surface temperature at Landsat resolution by fusing Landsat and MODIS data[J]. Remote Sensing of Environment, 2014, 145: 55-67.

[59] Zhang B, Zhang L, Xie D, et al. Application of Synthetic NDVI Time Series Blended from Landsat and MODIS Data for Grassland Biomass Estimation[J]. Remote Sensing, 2016, 8(1): 1-21.

[60] Zhu X, Chen J, Gao F, et al. An enhanced spatial and temporal adaptive reflectance fusion model for complex heterogeneous regions[J]. Remote Sensing of Environment, 2010, 114(11): 2610-2623.

[61] Cheng Q, Liu H, Shen H, et al. A Spatial and Temporal Nonlocal Filter-Based Data Fusion Method[J]. IEEE Transactions on Geoscience & Remote Sensing, 2017, PP(8): 1-13.

[62] Zhukov B, Oertel D, Lanzl F, et al. Unmixing-based multisensor multiresolution image fusion[J]. IEEE Transactions on Geoscience & Remote Sensing, 1999, 37(3): 1212-1226.

[63] Huang B, Zhang H. Spatio-temporal reflectance fusion via unmixing: accounting for both phenological and land-cover changes[J]. International Journal of Remote Sensing, 2014, 35(16): 6213-6233.

[64] Huang B, Song H. Spatiotemporal Reflectance Fusion via Sparse Representation[J]. IEEE Transactions on Geoscience & Remote Sensing, 2012, 50(10): 3707-3716.

[65] Song H, Huang B. Spatiotemporal Satellite Image Fusion Through One-Pair Image Learning[J]. IEEE Transactions on Geoscience and Remote Sensing, 2013, 51(4): 1883-1896.

[66] Dong C, Loy C C, He K, et al. Image Super-Resolution Using Deep Convolutional Networks[J]. IEEE Trans Pattern Anal Mach Intell, 2016, 38(2): 295-307.

[67] Gevaert C M, García-Haro F. A comparison of STARFM and an unmixing-based algorithm for Landsat and MODIS data fusion[J]. Remote Sensing of Environment, 2015, 156: 34-44.

[68] Zhu X, Helmer E, Gao F, et al. A flexible spatiotemporal method for fusing satellite images with different resolutions[J]. REMOTE SENS ENVIRON, 2016, 172: 165-177.

[69] Zhai H, Huang F, Qi H. Generating High Resolution LAI Based on a Modified FSDAF Model[J]. Remote Sensing, 2020, 12(1): 150-166.

[70] 王杰, 李卫朋. 基于灵活的时空融合模型的植被覆盖度与植被指数关系[J]. 草业科学, 2017, 34(2): 264-272.

[71] 王爽, 王承武, 张飞云. 基于FSDAF模型的干旱区典型绿洲城市夏季地表热岛效应时空演变研究[J]. 干旱区地理, 2021, 44(1): 118-130.

[72] Zhao F, Li Z, Verhoef W, et al. Simulation of solar-induced chlorophyll fluorescence by modeling radiative coupling between vegetation and atmosphere with WPS[J]. Remote Sensing of Environment, 2022, 277: 113075.

[73] Ja. S, Raissouni N. Toward remote sensing methods for land cover dynamic monitoring: application to Morocco[J]. International journal of remote sensing, 2000, 21(2): 353-366.

[74] 覃志豪, 李文娟, 徐斌, 等. 利用Landsat TM6反演地表温度所需地表辐射率参数的估计方法[J]. 海洋科学进展, 2004, 022: 129-137.

[75] Gao F, Masek J G, Schwaller M R, et al. On the Blending of the Landsat and MODIS Surface Reflectance: Predicting Daily Landsat Surface Reflectance[J]. IEEE Transactions on Geoscience and Remote Sensing, 2006, 44(8): 2207-2218.

[76] Zhu X, Helmer E, Gao F, et al. A flexible spatiotemporal method for fusing satellite images with different resolutions[J]. Remote Sensing of Environment An Interdisciplinary Journal, 2016, 172: 165-177.

[77] Guo M Z, Xiaolin. FSDAF 2.0: Improving the performance of retrieving land cover changes and preserving spatial details[J]. Remote Sensing of Environment: An Interdisciplinary Journal, 2020, 248(1): 223-238.

[78] Zhu X, Cai F, Tian J, et al. Spatiotemporal Fusion of Multisource Remote Sensing Data: Literature Survey, Taxonomy, Principles, Applications, and Future Directions[J]. Remote Sensing, 2018, 10(4): 527-550.

[79] 陈文裕, 夏丽华, 徐国良, 等. 2000—2020年珠江流域NDVI动态变化及影响因素研究[J]. 生态环境学报, 2022, 31(7): 1306-1316.

[80] 曹洁萍, 迟道才, 武立强, 等. Mann-Kendall检验方法在降水趋势分析中的应用研究[J]. 农业科技与装备, 2008, 1(5): 35-40.

[81] Mann H B. Nonparametric test against trend[J]. Econometrica, 1945, 13(3): 245-259.

[82] Kendall M G. Rank Correlation Methods[J]. British Journal of Psychology, 1990, 25(1): 86-91.

[83] Yan F, Wang Y J. Estimation of soil moisture from Ts-EVI feature space[J]. Acta Ecologica Sinica, 2009, 29(9): 4884-4891.

[84] 柳钦火, 辛景峰, 辛晓洲, 等. 基于地表温度和植被指数的农业干旱遥感监测方法[J]. 科技导报, 2007, 25(706): 12-18.

[85] 齐述华, 王长耀, 牛铮. 利用温度植被旱情指数(TVDI)进行全国旱情监测研究[J]. 遥感学报, 2003, 7(5): 420-427.

[86] 李树岩, 马志红. 河南省夏玉米生长季农业气候资源变化分析[J]. 河南农业科学, 2014, 41(7): 21-26.

[87] 吴领弟. 锡林郭勒地区植被物候遥感监测及其对气候变化的响应[D].呼和浩特: 内蒙古师范大学, 2018.

[88] 杨玲, 杨艳昭. 基于TVDI的西辽河流域土壤湿度时空格局及其影响因素[J]. 干旱区资源与环境, 2016, 30(2): 76-81.

[89] 吴孟泉, 崔伟宏, 李景刚. 温度植被干旱指数(TVDI)在复杂山区干旱监测的应用研究[J]. 干旱区地理, 2007, 30(1): 30-35.

[90] 蒋桂芹, 裴源生, 翟家齐. 农业干旱形成机制分析[J]. 灌溉排水学报, 2012, 31(6): 84-88.