本数据集为塔吉克斯坦国家级蝗虫生境风险监测栅格产品集，覆盖塔吉克斯坦全国范围，时间范围为2001—2022年，监测时段为每年2—7月，时间尺度为旬，空间分辨率为5 km。数据基于MODIS地表反射率、CHIRPS降水和ERA5-Land土壤湿度等多源遥感与再分析资料，在Google Earth Engine（GEE）平台完成数据获取、预处理、指标计算和风险产品生成，通过时间序列变化检测与历史同期对比，逐旬计算植被异常增长、孵化适宜条件、植被损伤及蝗灾爆发疑似风险，为蝗虫早期预警、孵化条件评估及农牧区植被损失评估提供数据支持。\n
本数据集基于多源遥感与再分析数据，在Google Earth Engine（GEE）平台完成预处理、指标计算和风险产品生成，覆盖2001—2022年每年2—7月，共形成22年×6个月×3旬=396期旬尺度监测产品。每个导出文件为单景GeoTIFF，包含5个风险波段，所有波段值均为整数类型，风险等级统一为：0=无风险/正常，1=低风险，2=中风险，3=高风险。\n
波段1：EARLY_WARNING（2月早期预警），基于NDVI较前旬异常快速增长，指示潜在越冬区蝗虫卵提前发育或植被返青异常，仅2月有效。\n
波段2：HATCH_RISK（3-4月孵化风险），耦合NDVI增长、正距平及表层土壤湿度（>0.06 m³/m³），指示土壤温湿条件适宜蝗卵孵化与若虫存活，仅3-4月有效。\n
波段3：DAMAGE_RISK（5-7月植被损伤风险），基于NDVI较历史同期显著下降且较前旬持续下滑，反映植被生物量受啃食胁迫，仅5-7月有效。\n
波段4：OUTBREAK_RISK（5-7月蝗灾爆发疑似），在损伤基础上增加“前两旬NDVI高值且持续加速下跌”条件并排除干旱区，用于锁定高概率爆发斑块，仅5-7月有效。\n波段5：LOCUST_RISK（综合蝗虫风险等级），取上述四个风险波段的逐像元最大值，代表该旬最终的综合风险警示级别，全年（2-7月）有效。", "ds_source": "
本数据集的4种原始数据源均通过Google Earth Engine（GEE）数据目录在线调用获取。包括NASA MODIS Terra MOD09A1 v061；UCSB-CHG CHIRPS Daily v2.0；ECMWF ERA5-Land Daily Aggregated；筛选研究区和日期范围获取0—7 cm表层土壤体积含水量；最后调用FAO/GAUL 2015 level 0行政区划数据获取塔吉克斯坦国境范围。", "ds_process_way": "
首先，对MOD09A1进行StateQA质量控制，剔除云、阴影及卷云等低质量观测，将反射率缩放至[0,1]并计算NDVI；随后按自然旬（1—10日、11—20日、21日至月末）对旬内有效NDVI取中值，生成当前旬标准影像。\n
其次，以当前旬为基准计算前旬、前两旬NDVI差值及下降速率，并在剔除目标年份的条件下，提取2001—2022年历史同期NDVI中值作为植被背景基准；同时计算历史同期月降水多年平均值、前30日累计降水及其相对历史同期均值的距平，以及前30日平均土壤湿度。\n
最后，根据蝗虫生活史阶段设置差异化判识逻辑：2月生成EARLY_WARNING，3—4月生成HATCH_RISK，5—7月生成DAMAGE_RISK和OUTBREAK_RISK，并逐像元取上述风险图层最大值生成LOCUST_RISK。所有中间及最终连续栅格统一采用适宜的空间聚合方式重采样至5 km，并统一投影至WGS84（EPSG:4326）， 最后按塔吉克斯坦国境范围裁剪并输出GeoTIFF。", "ds_quality": "
本数据集采用MOD09A1 v061、CHIRPS Daily和ERA5-Land等具有长期连续记录的标准遥感与再分析数据作为输入，并分别利用MODIS StateQA、ERA5-Land及原始数据质量信息开展有效观测筛选。旬尺度NDVI采用旬内中值合成，以降低云污染、短期噪声和单景异常值影响；降水和土壤湿度采用统一时间窗口计算，保证各指标时间对应关系。历史同期NDVI基准采用剔除目标年份后的多年数据计算，减少目标年份异常对基准的影响。风险产品按照明确的生活史阶段和指标阈值进行规则化计算，并对各风险图层进行统一空间尺度、投影和范围处理。", "ds_acq_start_time": "2001-02-01 00:00:00", "ds_acq_end_time": "2022-07-31 00:00:00", "ds_acq_place": "塔吉克斯坦", "ds_acq_lon_east": 75.23305555555555, "ds_acq_lat_south": 36.669444444444444, "ds_acq_lon_west": 67.51944444444445, "ds_acq_lat_north": 41.08305555555556, "ds_acq_alt_low": null, "ds_acq_alt_high": null, "ds_share_type": "open-access", "ds_total_size": 1732760, "ds_files_count": 377, "ds_format": "*.tif", "ds_space_res": "5km", "ds_time_res": "旬", "ds_coordinate": "无", "ds_projection": "", "ds_thumbnail": "56326785-563e-415b-90c4-b6643b57351e.png", "ds_thumb_from": 2, "ds_ref_way": "", "paper_ref_way": "", "ds_ref_instruction": "None", "ds_from_station": null, "organization_id": "b6ee8072-db07-4053-bcbc-78e22d72c5ec", "ds_serv_man": null, "ds_serv_phone": null, "ds_serv_mail": null, "doi_value": "", "subject_codes": [ "210" ], "quality_level": 0, "publish_time": "2026-10-14 15:20:12", "last_updated": "2026-10-14 15:20:12", "protected": false, "protected_to": "2027-09-05 00:00:00", "lang": "zh", "cstr": "11738.11.ncdc.db7841.2026", "i18n": { "en": { "title": "Tajikistan 5 km Dekadal-Scale Locust Monitoring Dataset for Key Locust Areas (2001–2022)", "ds_format": "*.tif", "ds_source": "
The four original data sources used in this dataset were all obtained through the online access to the Google Earth Engine (GEE) data catalog. These sources include NASA’s MODIS Terra MOD09A1 v061; UCSB-CHG’s CHIRPS Daily v2.0 data; and ECMWF’s ERA5-Land Daily Aggregated data. The data was filtered to only include information related to soil moisture at a depth of 0–7 cm, within specified study areas and date ranges. Finally, data from FAO/GAUL’s 2015 level 0 administrative division system was utilized to determine the boundaries of Tajikistan.", "ds_quality": "This dataset utilizes standard remote sensing and reanalysis data such as MOD09A1 v061, CHIRPS Daily, and ERA5-Land, which provide continuous monitoring over long periods of time. Additionally, techniques like MODIS StateQA, ERA5-Land, and raw data quality information are employed to perform effective filtering of observations. The ten-day NDVI values are calculated using the median values within each ten-day period, thereby reducing the impact of cloud pollution, short-term noise, and outliers from individual images. Precipitation and soil moisture data are calculated using a unified time window, ensuring that the time relationships between various indicators are maintained. The baseline NDVI values for historical periods are determined by calculating values from multiple years after excluding the target year, thereby minimizing the influence of anomalies in the target year on the baseline values. Risk products are calculated based on specific life stages and indicator thresholds. Finally, all risk layers are processed to ensure consistent spatial scale, projection, and range settings.", "ds_ref_way": "", "ds_abstract": "This dataset contains information on the risk levels of locust habitats in Tajikistan. It covers the entire territory of Tajikistan, with data collected from 2001 to 2022. The monitoring period is from February to July each year, and the time scale used is the ten-day period. The spatial resolution is 5 kilometers. The data is derived from multiple sources, including MODIS surface reflectivity, CHIRPS precipitation data, and ERA5-Land soil moisture data. Data acquisition, preprocessing, calculation of indicators, and generation of risk products are all performed on the Google Earth Engine platform. By analyzing trends over time and comparing with historical data, the dataset provides information on abnormal vegetation growth, suitable conditions for locust hatching, damage to vegetation, and suspected risks of locust outbreaks. This data supports early warnings regarding locust infestations, assessment of hatching conditions, and evaluation of vegetation losses in agricultural areas.\r\n\r\nThe dataset is based on multi-source remote sensing and reanalysis data. Preprocessing, indicator calculation, and risk product generation are carried out on the Google Earth Engine platform. Data covers the years from 2001 to 2022, with monitoring periods from February to July each year. In total, there are 22 years × 6 months × 3 ten-day periods, resulting in 396 monitoring products. Each exported file is a single-image GeoTIFF format, containing 5 risk bands. All band values are integer types. The risk levels are defined as follows: 0 = no risk/normal, 1 = low risk, 2 = moderate risk, 3 = high risk.\r\n\r\nBand 1: EARLY_WARNING (early warning in February). This band indicates potential early development of locust eggs or abnormal greening of vegetation, based on rapid growth of NDVI compared to the previous ten-day period. This band is only valid in February.\r\n\r\nBand 2: HATCH_RISK (hatching risk from March to April). This band combines NDVI growth, positive anomalies, and surface soil moisture levels (>0.06 m³/m³) to indicate suitable conditions for locust eggs to hatch and nymphs to survive. This band is only valid from March to April.\r\n\r\nBand 3: DAMAGE_RISK (vegetation damage risk from May to July). This band reflects significant decline in NDVI compared to historical levels, along with continuous downward trends compared to the previous ten-day period. This indicates that vegetation biomass has been damaged by herbivores. This band is only valid from May to July.\r\n\r\nBand 4: OUTBREAK_RISK (suspected locust outbreaks from May to July). This band adds the condition of high NDVI values in the first two ten-day periods, along with continuous declining values, to identify high-probability outbreak areas. This band is only valid from May to July.\r\n\r\nBand 5: LOCUST_RISK (comprehensive locust risk level). This band takes the maximum value of the four risk bands mentioned above per pixel, representing the overall risk level for that ten-day period. This band is valid throughout the year, from February to July.", "ds_time_res": "Xun", "ds_acq_place": "Tajikistan", "ds_space_res": "5km", "ds_projection": "", "ds_process_way": "
Firstly, perform StateQA quality control on MOD09A1, removing low-quality observations such as clouds, shadows, and cirrus clouds. Then, scale the reflectivity values to the range [0,1] and calculate the NDVI value. Next, take the median of the valid NDVI values for each natural ten-day period (from day 1 to day 10, from day 11 to day 20, and from day 21 until the end of the month), to generate standard images for that period.\r\n
Secondly, using the current ten-day period as a reference, calculate the differences and decline rates of NDVI values for the previous ten days and the first two ten-day periods. After excluding data from the target year, extract the median NDVI values for the same period from 2001 to 2022, to serve as a baseline for vegetation conditions. Additionally, calculate the annual average monthly precipitation levels, the cumulative precipitation over the past 30 days, and their deviation from the historical average. Also, determine the average soil moisture level over the past 30 days.\r\n
Finally, establish differentiated recognition logic based on the different stages of locust life cycles: generate EARLY_WARNING for February, HATCH_RISK for March–April, DAMAGE_RISK and OUTBREAK_RISK for May–July. For each pixel, take the maximum value from these risk layers to generate LOCUST_RISK. All intermediate and final raster data are resampled using appropriate spatial aggregation methods to a resolution of 5 kilometers. They are then projected onto the WGS84 coordinate system (EPSG:4326). Finally, crop the data according to the borders of Tajikistan and save them in GeoTIFF format.
", "ds_ref_instruction": "" } }, "submit_center_id": "ncdc", "data_level": 0, "recommendation_value": 0, "license_type": "https://creativecommons.org/licenses/by/4.0/", "doi_reg_from": "reg_local", "cstr_reg_from": "reg_local", "doi_not_reg_reason": null, "cstr_not_reg_reason": null, "is_paper_in_submitting": false, "belong_to_nieer": false, "allow_update_data": false, "created_from": "xmhj", "ds_topic_tags": [ "蝗虫风险监测" ], "ds_subject_tags": [ "农学" ], "ds_class_tags": [], "ds_locus_tags": [ "塔吉克斯坦" ], "ds_time_tags": [ 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021, 2022 ], "ds_contributors": [ { "true_name": "钱静", "email": "jing.qian@siat.ac.cn", "work_for": "中国科学院深圳先进技术研究院", "country": "中国" }, { "true_name": "孙博", "email": "sunbo@siat.ac.cn", "work_for": "中国科学院深圳先进技术研究院", "country": "中国" }, { "true_name": "魏树杰", "email": "sj.wei@siat.ac.cn", "work_for": "中国科学院深圳先进技术研究院", "country": "中国" }, { "true_name": "邢秀为", "email": "xw.xing@siat.ac.cn", "work_for": "中国科学院深圳先进技术研究院", "country": "中国" } ], "ds_meta_authors": [ { "true_name": "魏树杰", "email": "sj.wei@siat.ac.cn", "work_for": "中国科学院深圳先进技术研究院", "country": "中国" } ], "ds_managers": [ { "true_name": "魏树杰", "email": "sj.wei@siat.ac.cn", "work_for": "中国科学院深圳先进技术研究院", "country": "中国" } ], "category": "生态" }