GLOBAL MODIS NDVI/LAI and NOAA AVHRR GIMMS NDVI datasets

MODIS NDVI data of the globe 2000 - 2019 Data delivered by Boston University and further prepared as derivates by Kjell Arild Høgda (NORCE) and Hans Tømmervik (NINA).   Derivates of the Boston data: MODIS_MaxNDVI_Annual: Maximum NDVI for each year in the period 2000-2019 (20 years). MODIS_MaxNDVI_Mean: Maximum NDVI averaged over the period 2000-2019.  MODIS_MaxNDVI_Median: Median for Maximum NDVI over the years 2000-2019.  MODIS_MaxNDVI_Percent Change: Percent Change of Maximum NDVI over the period 2000-2019. MODIS_MaxNDVI_Percent Deviation_Mean: Mean of Percent Deviation_ of Maximum NDVI for each year in the period 2000-2019. MODIS_MaxNDVI_Percent Deviation_Median: Median of Percent Deviation of Maximum NDVI for each year in the period 2000-2019 MODIS_MaxNDVI_Trend: Linear trend of MaxNDVI over the period 2000-2019. MODIS_TI_NDVI_Annual: MODIS Time integrated NDVI for each year in the period 2000-2019. MODIS_TI_NDVI_Mean: MODIS Time integrated NDVI averaged over the period 2000-2019. MODIS_TI_NDVI_Median: Median MODIS Time Integrated NDVI averaged for the period 2000-2019. MODIS_TI_NDVI_Percent_Change: Percent Change of Time Integrated NDVI over the period 2000-2019. MODIS_TI_NDVI_Percent_Dev_Mean: Mean of Percent Deviation of Time integrated NDVI for each year in the period 2000-2019. MODIS_TI_NDVI_Percent_Dev_Median: Median of Percent Deviation of Time integrated NDVI for each year in the period 2000-2019. MODIS_TI_NDVI_Trend: Linear trend of Time Integrated NDVI over the period 2000-2019.     MODIS LAI - Leaf Area Index of the globe 2000 - 2019 Data delivered by Chi Chen (Boston University) and further prepared as derivates by Kjell Arild Høgda (NORCE) and Hans Tømmervik (NINA).  Derivates of the Boston data: MODIS_MaxLAI_Annual: Maximum Leaf Area Index for each year in the period 2000-2019 (20 years). MODIS_MaxLAI_Mean: Maximum Leaf Area Index averaged over the period 2000-2019.  MODIS_MaxLAI_Percent Change: Percent Change of Maximum Leaf Area Index over the period 2000-2019. MODIS_MaxLAI_Trend: Linear trend of MaxLAI over the period 2000-2019.   NOAAAVHRR GIMMS NDVI ncdisp(‘ndvi3g_geo_v1_2015_0712.nc4’) approximate size of each netcdf4 file: 448MB Source:           ndvi3g_geo_v1_2015_0712.nc4 Format:            netcdf4_classic Global Attributes:            FileName        = 'ndvi3g_geo_v1_2015_0712.nc4'            Institution       = 'NASA/GSFC GIMMS'            Data               = 'NDVI3g version 1'            Reference      = '1. Pinzon, J.E.; Tucker, C.J.                                                                          A Non-Stationary 1981-2012 AVHRR NDVI3g Time Series.                                                          Remote Sens. 2014, 6, 6929-6960.                                                                        2. Pinzon, J.E.; Tucker, C.J.                                                                                    A Non-Stationary 1981-2015 AVHRR NDVI3g.v1 Time Series: an update.                                             Remote Sens. 2016, in preparation’                                '            Comments Version1      = 'version1 includes two major fixes (a and b), and three minor (c-g):               (a) Reprocessed Level 2 entire SeaWIFS mission for the land products                     to reduce artifacts in the data, particularly changes in calibration                     after 2006 that generates drops in ndvi lower values.                     OB.DAAC / Ocean Biology Processing group NASA/GSFC 616  (april 2016)               (b)  Recovered ndvi negative values of snow-covered regions in winter                     Northern latitudes.  In Version0, we masked them with zero values,                     creating artifacts in phenology parameters.               (c)  Arranged data in ncd format, compiled it in two nc4 files a year.                     Each nc4 file includes 6 months of ndvi data (jan-jun and jul-dec),                     with a total of 12 (15-day) composites each semester.               (d)  Rescaled ndvi values and splitted the flag values from them.                (e)  Added a new variable, percentile, to represent the distribution of                     ndvi values in the time series. Range 10*[0, 100]                (f)  Flag values are (simpler):                                               flag 0: ndvi without apparent issues (good value)                                               flag 1: ndvi retrieved from spline interpolation                                               flag 2: ndvi retrieved from seasonal profile (possible snow/cloud)                 (g) Flag values are embeded on the percentile variable:  2000*flag + percentile.                     Thus, the actual percentile three ranges [0 1000], [2000 3000] and [4000 5000]                       could provide direct nformation of how interpolation is affecting the time series.      '            Temporalrange         = '1981-07-01 -> 2015-12-31'            Year                          = 2015            RangeSemester       = 'Jul 1 - Dec 31 (7:0.5:12.5)'            SpatialResolution     = '1/12 x 1/12 degrees'            TemporalResolution   = '1/24 a year'            _fill_val                       = -32768            NorthernmostLatitude = '90'            SouthernmostLatitude = '-90'            WesternmostLongitude = '-180'            EasternmostLongitude = '180' Dimensions:            lon  = 4320            lat  = 2160            time = 12 Variables:     lon                  Size:       4320x1            Dimensions: lon            Datatype:   double     lat                  Size:       2160x1            Dimensions: lat            Datatype:   double     time                 Size:       12x1            Dimensions: time            Datatype:   double     satellites            Size:       12x1            Dimensions: time            Datatype:   int16     ndvi                 Size:       4320x2160x12            Dimensions: lon,lat,time            Datatype:   int16            Attributes:                        units         = '1'                        scale         = 'x 10000'                        missing_value = -5000                        valid_range   = [-0.3           1]     percentile            Size:       4320x2160x12            Dimensions: lon,lat,time            Datatype:   int16            Attributes:                        units       = '%'                        scale       = 'x 10'                        flags       = 'flag 0: from data                flag 1: spline interpolation  flag 2: possible snow/cloud cover'                        valid_range = 'flag*2000 + [0 1000]'