Seasonal maximum vegetation condition index (VCIx) is a remote sensing- based indicator introduced by CropWatch in 2014 for crop growth condition monitoring. VCIx adopts the general concept of Vegetation Condition Index (VCI) but stretches the length of temporal observation window from a short time slot, fixed by satellite sensor, to a period that can reflect various crop growth stages (crop phenology). In this way, it reduces the uncertainty of remote sensing index-based crop condition monitoring caused by inter-annual shifts (delay or advance) of crop phenology over different years. In CropWatch, VCIx is presented as a raster map at global extent with 1 Km resolution, updated every three months. Pixel values usually fall between 0 and 1. Based on the VCIx values, crop growth condition can be categorized into four levels: Level 1: VCIx＜0.5, indicating poor crop growth condition which is below the average of the previous 5 years (5YA) and 0 means as bad as the worst recent year; Level 2: 0.5≤VCIx＜0.8, indicating slight above 5YA situation; Level 3: 0.8≤VCIx≤1.0, indicating that crop condition is better than the 5YA but below the optimal condition during the previous five years, 1 means as good as the best recent year. Level 4: VCIx＞1.0, indicating a new record level of crop growth condition which exceeds the optimal condition of the previous 5 years. VCIx is calculated based on NDVI time series (MODIS). Peak NDVI during the monitoring period is compared with the historic (previous five years) minimum NDVI during the same period and normalized by the historical range of NDVI values for the same period. As NDVI values may be distorted by cloud or non-vegetation pixels, an empirical minimum vegetation NDVI value (0.15) is introduced in VCIx computation. In case the minimum NDVI of the monitoring period is lower than the empirical value (0.15), the empirical value (0.15) is used in the computation. Considering the genetic development and improvement of crops seeds, crops at monitoring year are hardly comparable with the same ones cultivated ten years ago. CropWatch uses previous five years, instead of a longer period, as the reference period when deriving the historic agronomic indicators. Detailed documentation on VCIx can be found at: http://cprs.patentstar.com.cn/Search/Detail?ANE=4CAA9DHB9DFABDIA9ICC9IGFAIIA9FFDCICA5CAA9ICC9DEB

Cropped arable land fraction (CALF) was introduced to demonstrate the proportion of cropped arable land to the total arable land over a certain geographic area (Major Production Zones (MPZs), countries or sub-national units). Monitoring the dynamic changes in arable land utilization, specifically the dynamic identification of cropped and uncropped arable land, is important. CALF can reflect the rotation pattern of different crops and the change of cultivated land planting intensity, especially for early warning of crop planting area. On the basis of an analysis of profiles of time series NDVI, Savitzky-Golay filters are used to smooth the noise in NDVI curves, and Lagrange polynomials are employed to extract the extreme points for the smoothed NDVI curves. A threshold method associated with NDVI curve analysis is used to identify dynamic changes in the distribution of cropped and uncropped arable land. CALF over those regions was then calculated based on cropped and uncropped map and zonal statistical analysis. In CropWatch, CALF is presented as a statistical value updated every three months from raster map at global extent with 1 Km resolution for each spatial unit derived. The statistical value reflects the overall planting ratio. The Global raster maps show an area as cropped if at least one of the remote sensing observations during the monitoring period is categorized as "cropped". Uncropped means that no crops were detected over the whole reporting period. Based on the number of pixels for marked as "cropped" or "uncropped" within a certain spatial unit, CALF value is derived by the proportion of cropped pixels to the total arable land pixels (or cropped + uncropped pixels). CALF values are compared to the average value for the previous five years, with departures expressed in percentage. CALF is used as an early warning indicator for the planted area at the period of one month after emergence. Considering the genetic development and improvement of crops seeds, crops at monitoring year are hardly comparable with the same ones cultivated ten years ago. CropWatch uses previous five years, instead of a longer period, as the reference period when deriving the historic agronomic indicators. Detailed documentation on CALF can be found at: http://www.cropwatch.com.cn/htm/en/files/201682105626480.pdf

Cropping intensity, defined as the number of cropping cycle(s) per year, is an important indicator to measure arable land use intensity. Tracking the change in cropping intensity can help assess the past development of the food production system and inform future agro-policies. All available images of top-of-atmosphere (TOA) reflectance from Landsat-7 ETM+, Landsat-8 OLI, Sentinel-2 MSI and MODIS during 2016–2018 were used for cropping intensity mapping via the GEE platform. To overcome the multi-sensor mismatch issue, an inter-calibration approach was adopted, which converted Sentinel-2 MSI and Landsat-8 OLI TOA reflectance data to the Landsat-7 ETM+ standard. Then the calibrated images were used to composite the 16-day TOA reflectance time series based on maximum composition method. To ensure data continuity, the MODIS NDVI product was used to fill temporal gaps with the following steps. First, the 250-m MODIS NDVI product was re-sized to 30-m using the bicubic algorithm. Then, the Whittaker algorithm was applied to the gap filled NDVI time series to smooth the NDVI time series. Two phenology metrics were introduced, mid-greenup and mid-greendown, which were derived as the day of year (DOY) at the transition points in the greenup and greendown periods when the smoothed NDVI time series passes 50% of the NDVI amplitude. An interval starting from mid-greenup and ending at mid-greendown is defined as a growing phenophase, and an interval moving from mid-greendown to mid-greenup a non-growing phenophase. Based on this phenophase-based approach, the global cropping intensity at 30m resolution (GCI30) was mapped. The results were validated based on a large number of ground-based samples obtained using GVG (GPS, Video and GIS) smart phone application and other crowd-sourcing dataset. The global cropping intensity dataset at 30m includes two layers. The first layer indicates the average cropping intensity during the three years from 2016 to 2018 with noData value or masked areas assigned to -1. The valid values for the first layer are 1, 2, and 3 representing single cropping, double cropping or triple cropping. The second layer keeps the original total number of crop cycles from 2016 to 2018 with noData value or masked areas assigned to -1. Continuous cropping or number of crop cycles larger than 3 per year are indicated with value of 127. Detailed documentation on the methodology of GCI30 can be found at the following two published papers: https://www.sciencedirect.com/science/article/abs/pii/S0034425720304685 https://essd.copernicus.org/articles/13/4799/2021/