Data Products

Learn about our process for data product development, see a full list of our data products as well as information on our models and software in our Data Management Plan.

Recently Released Gridded Data Products

The following gridded data products are currently offered by the USA National Phenology Network, but have not yet undergone the USGS Fundamental Science Practices review. As such, they are considered "Provisional." For a complete list of gridded datap products, see the USA National Phenology Network Gridded Product Documentation.

1.0 Accumulated Growing Degree Day Products

Heat accumulation is commonly used as a way of predicting the timing of phenological transitions in plants and animals, including when plants exhibit leaf out, flowering, or fruit ripening, or when insects emerge from dormancy (Cross and Zuber, 1972; McMaster and Wilhelm, 1997). This is typically expressed as accumulated heat units, either Growing Degree Hours or Growing Degree Days. Growing degree day thresholds have been established for many species, and are commonly used in agriculture, horticulture, and pest management to schedule activities such as harvesting, pesticide treatment, and flower collection.

1.1 Daily Contemporary and Forecasted Temperature Accumulations for Alaska

Underlying climate data: National Oceanic and Atmospheric Administration (NOAA) National Centers for Environmental Prediction (NCEP) Real-Time Mesoscale Analysis (RTMA) products, NOAA NCEP Unrestricted Mesoscale Analysis (URMA) products, and NOAA National Weather Service (NWS) National Digital Forecast Database (NDFD)

• Spatial resolution: 3 kilometer (km)
• Spatial extent: Mainland Alaska
• Temporal resolution: Daily
• Temporal extent: Beginning January 1, 2017, January 1 through current day plus 1- to 6-day forecasts
• Units: Accumulated growing degree days (AGDD) based on January 1 start date and base temperature of 32 °F or 50 °F
• Additional features available: State boundaries

Description: These layers represent the amount of heat accumulated in each pixel, for Alaska. There is one layer for each day of the year up through the current day, plus 6 days into the future. Grid cell values begin at 0 and increase as heat units begin to accumulate. These grids are updated on a nightly process.  By adjusting the day of year that is viewed, the user can explore the amount of heat accumulated at any location on that day. The user may also explore the growing degree days anticipated to accumulate in the next 1 to 6 days, based on short-term temperature forecast products (NDFD). For more information on workflow and stability of forecasted data see Crimmins et al., 2017, Section 2.2.1.

2.0 Extended Spring Indices Products

The Extended Spring Indices are mathematical models that predict the “start of spring” (timing of first leaf or first bloom) at a particular location. These models were constructed using historical observations of the timing of first leaf and first bloom in a cloned lilac cultivar (Syringa X chinensis ‘Red Rothomagensis’) and two cloned honeysuckle cultivars (Lonicera tatarica L. ‘Arnold Red’ and Lonicera korolkowii Stapf, also known as ‘Zabelii’). The model outputs are first leaf and first bloom date for a given location. Model output is available for each of these species individually, or as an average of the three species.

2.1 Historical Spring Indices (BEST)

• Underlying climate data: Berkeley Earth Surface Temperature (BEST)
• Spatial resolution: 1° latitude by 1 °  longitude
• Spatial extent: Northwestern semi-sphere, including CONUS, AK, and HI
• Temporal resolution: Yearly
• Temporal extent: 1880-2013
• Units: Day of year
• Additional features available: State boundaries (composite maps only)

Description: These layers are annual representations of the days of year that the requirements for the first leaf or first bloom Spring Indices were met, from 1880 to 2013 (one layer per year), calculated using BEST daily minimum and maximum temperature data (Tmin and Tmax, respectively).  These layers are not available for individual species threshold models, only the Spring Indices (the average of the three species threshold models). 

2.2 Historical Spring Indices (NCEP)

• Underlying climate data: NOAA NCEP URMA products
• Spatial resolution: 2.5 km
• Spatial extent: Contiguous United States
• Temporal resolution: Yearly
• Temporal extent: 2016 through the prior calendar year
• Units: Day of year
• Additional features available: State boundaries (composite maps only)

Description: These layers are annual representations of the days of year that the requirements for the first leaf or first bloom Spring Indices were met, from 2016 through the prior calendar year (one layer per year), calculated using NOAA NCEP URMA daily minimum and maximum temperature data (Tmin and Tmax, respectively). For each year, both first leaf and first bloom layers are available for each of the three species threshold models as well as for the first leaf and first bloom Spring Indices, which are the average of the three first leaf or first bloom species threshold models. These layers are available from 2016 forward, to enable inter-comparison with PRISM products, and to provide continuity in data availability. 

2.3 Historical Spring Index Anomalies (NCEP)

• Underlying climate data: NOAA NCEP URMA products
• Spatial resolution: 2.5 km
• Spatial extent: Contiguous United States
• Temporal resolution: Yearly
• Temporal extent: 2016 through the prior calendar year
• Units: Days
• Additional features available: State boundaries (composite maps only)

Description: These layers show the difference, in days, between the Historical Spring Indices (NCEP) and the 30-Year Average for Spring Indices. The layers show how advanced or lagged the day of year of the first leaf and first bloom Spring Indices are, for the current year in a given location, compared to long-term averages. These values are calculated as follows:

(Current year’s values) – (30-year average values)

such that negative values represent locations that have reached the SI-x requirements earlier than average, and positive values represent locations that have reached the SI-x requirements later than average. These layers are not available for individual species threshold models, only the Spring Indices (the average of the three species threshold models). These layers are available from 2016 forward, to provide continuity in data availability.

2.4 Daily Contemporary and Short-term Forecasted Spring Indices for Alaska

• Underlying climate data: NOAA NCEP RTMA products, NOAA NCEP URMA products, and NOAA NWS NDFD
• Spatial resolution: 3 km
• Spatial Extent: Mainland Alaska
• Temporal resolution: Daily
• Temporal extent: Beginning January 1, 2017, January 1 through current day plus 1- to 6-day forecasts
• Units: Day of year
• Additional features available: State boundaries

Description: These layers represent the progression of the first leaf and first bloom Spring Indices in the current year, for Alaska. There is one layer for each day of the year up through the current day, plus 6 days into future. Both first leaf and first bloom layers are available for each of the three species threshold models as well as for the first leaf and first bloom Spring Indices, which are the average of the three first leaf or first bloom species threshold models.

These grids are updated on a nightly process. A cell is not shaded with the day of year that the Spring Index requirements are met until the requirements for each of the three species threshold models have been met. By adjusting the day of year that is viewed, the user can explore the locations where the Spring Index requirements have been reached, as well as pixels where the Spring Index requirements are anticipated to be met in the coming 6 days, based on short-term temperature forecast products (NDFD). For more information on workflow and stability of forecasted data see Crimmins et al., 2017, Section 2.2.2.

2.5 Historical Spring Indices for Alaska (NCEP)

• Underlying climate data: NOAA NCEP RTMA products, NOAA NCEP URMA products, and NOAA NWS NDFD
• Spatial resolution: 3 km
• Spatial Extent: Mainland Alaska
• Temporal resolution: Yearly
• Temporal extent: 2017 through the prior calendar year
• Units: Day of year

Description: These layers are annual representations of the days of year that the requirements for the first leaf or first bloom Spring Indices were met, for Alaska, from 2017 through the prior calendar year (one layer per year), calculated using NOAA NCEP URMA daily minimum and maximum temperature data (Tmin and Tmax, respectively). For each year, both first leaf and first bloom layers are available for each of the three species threshold models as well as for the first leaf and first bloom Spring Indices, which are the average of the three first leaf or first bloom species threshold models. These layers are available from 2017 forward, to enable inter-comparison with PRISM products, and to provide continuity in data availability. 

2.6 30-Year Average for Extended Spring Indices

• Underlying climate data: PRISM
• Spatial resolution: 4 km
• Spatial extent: Contiguous United States
• Temporal resolution: Not applicable
• Temporal extent: 1981-2010 Average
• Units: Day of year
• Additional features available: State boundaries 

These two layers show the 30-year average day of year that the requirements were met for the first leaf and first bloom Spring Indices. These long-term averages were created by averaging the annual Spring Index layers generated for each year in the span 1981 - 2010. These grids were generated using historical PRISM data. 

2.7 Historical Annual Spring Index Anomaly (PRISM)

• Underlying climate data: PRISM
• Spatial resolution: 4 km
• Spatial extent: Contiguous United States
• Temporal resolution: Yearly
• Temporal extent: 1981 to previous calendar year
• Units: Days
• Additional features available: State boundaries 

These two layers show the difference, in days, between the Historical Annual Spring Index and the 30-Year Average Spring Index at 4km resolution for first leaf and first bloom, calculated using PRISM data. The layer shows how advanced or lagged the day of year of the first leaf index is for each year, compared to the long-term average day of year. The Extended Spring Indices are models that predict the "start of spring" (timing of leaf out or bloom) at a particular location. These values are calculated as follows:

(Current year’s values) – (30-year average values)

such that negative values represent locations that have reached the SI-x requirements earlier than average, and positive values represent locations that have reached the SI-x requirements later than average. 

2.8 Spring Index Return Intervals

Source data: Daily Spring Index Anomalies (2.2.4)
Spatial resolution: 4 km
Temporal resolution: Yearly
Temporal extent: 2018-present
Units: Return interval (years)
Additional features available: None
Available Via: Images on Status of Spring Page (current year); Images on Spring Index technical page (planned)

Description: The Spring Index Return Intervals for leaf out and bloom represent the frequency with which spring is as late or early as it is in the current year. It serves to contextualize the current year’s spring onset anomaly. For example, if the Daily Anomaly product shows leaf out 20 days earlier than average, for a given location, the Return Interval product shows how often that location has experienced leaf out at least 20 days early. To create this product, the current year Spring Index Anomaly value is compared to the anomaly value for each year from 1981 through the prior year. The frequency with which a spring was at least this early (or late) is calculated by taking the number of years in the record divided by the count of years that were earlier (or later) than the current year.

3.0 Temperature Input Products - Daily Minimum and Maximum Temperatures

The daily minimum and maximum temperature data layers used to produce the Accumulated Growing Degree Days and Extended Spring Indices products described above are also made available for access and analysis.

3.1 Daily Minimum and Maximum Temperatures for Alaska

• Underlying climate data: NOAA NCEP RTMA products, NOAA NCEP URMA products, and NOAA NWS NDFD
• Spatial resolution: 3 km
• Spatial extent: Mainland Alaska
• Temporal resolution: Daily
• Temporal extent: Beginning January 1, 2016, January 1 through current day plus 1- to 6-day forecasts
• Units: Degrees Fahrenheit

Description: These layers represent daily temperature minimums and maximums as expressed in degrees Fahrenheit in each cell for Alaska. There are two grids for each calendar day—one for minimum temperatures, and one for maximum temperatures. By adjusting the day of year that is viewed, the user can explore minimum and maximum temperatures reported for that day. The user may also explore the anticipated minimum and maximum temperatures for the coming 1 to 6 days, based on short-term temperature forecast products (NDFD). For more information on workflow, format and stability of forecasted data see Crimmins et al., 2017, Section 2.3.

4.0 Pheno Forecast Products

4.1 Pest Pheno Forecasts

Source data: Daily Accumulated Growing Degree Days
Spatial resolution: 2.5 km
Temporal resolution: Daily
Temporal extent: 2018-present
Units: Growing Degree Days
Additional features available: Limited to known distribution of pest species by state
Available Via: Images (current year); Viz Tool (2018-current year).

Description: Pheno Forecast maps predict key life cycle stages in invasive and pest species to improve management efficacy. For insect pest species, Pheno Forecasts are based on published growing degree day (GDD) thresholds for key points in species life cycles. These key points typically represent life cycle stages when management actions are most effective. These maps are updated daily and available 6 days in the future. Species-specific threshold values, start dates, and base temperatures can be found in Crimmins et al. 2020.

Crimmins, T. M., K. L. Gerst, D. G. Huerta, R. L. Marsh, E. E. Posthumus, A. H. Rosemartin, J. Switzer, J. F. Weltzin, L. Coop, N. Dietschler, D. A. Herms, S. Limbu, R. T. Trotter, III, and M. Whitmore. 2020. Short-Term Forecasts of Insect Phenology Inform Pest Management. Annals of the Entomological Society of America.

4.2 Invasive Buffelgrass Pheno Forecast

Source data: Daily Accumulated Precipitation, PRISM
Spatial resolution: 4 km
Temporal resolution: Daily
Temporal extent: 2019-present
Units: Accumulated Precipitation (inches)
Additional features available: Limited to Arizona
Available Via: Geoserver Request Builder, Viz Tool

Description: Pheno Forecast maps predict key life cycle stages in invasive and pest species, to improve management efficacy. The buffelgrass Pheno Forecast is based on known precipitation thresholds for triggering green-up to a level where management actions are most effective based on research presented in Wallace et al. 2016. Daily values represent the 24 day prior precipitation accumulation. 24 day accumulated precipitation values are derived from PRISM daily precipitation data. These maps are updated daily and predict green-up one to two weeks in the future.

PRISM Climate Group, Oregon State University, http://prism.oregonstate.edu, created 4 Feb 2004.

Wallace, C.S.A.; Walker, J.J.; Skirvin, S.M.; Patrick-Birdwell, C.; Weltzin, J.F.; Raichle, H. 2016. Mapping Presence and Predicting Phenological Status of Invasive Buffelgrass in Southern Arizona Using MODIS, Climate and Citizen Science Observation Data. Remote Sens. 8: 524.

4.3 Winter Wheat Development Pheno Forecast

Source data: NOAA NCEP RTMA products, NOAA NCEP URMA products, and NOAA NWS NDFD
Spatial resolution: 2.5 km
Temporal resolution: Daily
Units: GDD
Available via: Images (current year); Geoserver Request Builder; Viz tool 

Description: The USA-NPN winter wheat development forecast predicts the developmental stage of winter wheat from tillering through seed development in real-time. The USA-NPN winter wheat development forecast is based on the CERES-wheat model (Ritchie 1991), which predicts developmental stage using temperature, sun angle, and varietal inputs. The USA-NPN winter wheat development forecast approximates across varieties to provide a general forecast for winter wheat development and may reflect developmental stage in some varieties more closely than others.

Winter wheat developmental stages are estimated as a function of accumulated growing degree days (GDDs), following Ritchie (1991). Daily GDD accumulations are calculated using a Jan 1 start date, 32F base temperature, 78.8F upper threshold, and the simple averaging method. Emergence is assumed to have happened in the fall. Accordingly, accumulations in this forecast begin from 0 at the start of tillering (equivalent to  1265 GDDs as described in Ritchie 1991).  Daily GDD accumulations are penalized based on whether vernalization requirements have been met and the number of hours of daylight at that location. All calculations and values are in Farenheit. 

To determine the vernalization penalty for each pixel, vernalization days are calculated beginning on Oct 1 as follows. If the daily average temperature (Tavg) falls between 32-44.6F, the day is considered a full VD. If the daily average temperature falls between 44.6-64.4F, a fraction of a VD is calculated using the formula (approximated from Ritchie 1991, Fig 3-1): [Fraction VD = -0.06(Tavg) + 3.38]. The vernalization penalty for a location on a given day is a function of VDs, calculated using the formula (approximated from Ritchie 1991, Fig 3-2): [VDPenalty = 0.015(VDs) + 0.21]. Once a pixel reaches 50VDs, the vernalization penalty is no longer calculated. The photoperiod penalty for each pixel is calculated using the formula (approximated from Ritchie 1991, Fig 3-3): [PhotoperiodPenalty =  0.075(hours daylight) - 0.24]. The GDD accumulation for each pixel is penalized (reduced) for each pixel by multiplying it by either the VDPenalty or the PhotoperiodPenalty, whichever value is smaller.

Ritchie, J.T. 1991. Wheat phasic development. p.31-54. In Hanks and Ritchie (ed.) Modeling Plant and Soil Systems. Agronomy Monograph 31, ASA, CSSSA, SSSA, Madison, WI.

5.0 Land Surface Phenology Products

The MODIS Land Cover Dynamics Product (MLCD) provides global land surface phenology (LSP) data from 2001-present. MLCD serves a wide variety of applications and is currently the only source of operationally produced global LSP data. MLCD data have enabled important discoveries about the role of climate in driving seasonal vegetation changes, helped to create improved maps of land cover, and support ecosystem modeling efforts, among many other important applications. The LSP Climate Indicators (LSP-CI) dataset is a curated collection of the most relevant phenological indicators: a measure of spring and autumn timing and a measure of seasonal productivity. Statistically robust estimates of long-term normals (median and median absolute deviation, MAD), significance-screened trends (Theil-Sen slope magnitude where p<=0.05), and interannual anomalies (in days as well as multiples of MAD) have been computed for these three phenological indicators. The data have been mosaiced across CONUS, reprojected and resampled to a more familiar spatial reference system that matches complementary datasets and delivered in the universally accessible GeoTIFF format.

5.1 Mid Greenup Median

Source data: MODIS MCD12Q2 v006
Spatial resolution: 2.5 km
Temporal resolution: 17 year median
Temporal extent: 2001-2017
Units: Day of Year
Additional Features: none
Available Via: Images (LSP landing page); Geoserver Request Builder

Description: The day of year that half of peak greenness occurs is considered an important metric of spring leaf-out. Greenness is calculated from satellite sensors that measure light reflectance from vegetation. Pixel values are the median day of year (2001-2017) the greenness of a pixel first reached 50% of the annual maximum. Negative values indicate that the greenup started prior to Jan 1 and the date of maximum greenness occurred after Jan 1.

5.2 Mid Greendown Median

Source data: MODIS MCD12Q2 v006
Spatial resolution: 2.5 km
Temporal resolution: 17 year median
Temporal extent: 2001-2017
Units: Day of Year
Additional Features: none
Available Via: Images (LSP landing page); Geoserver Request Builder

Description: The day of year half-way between peak greenness and senescence is considered an important metric of autumn leaf color change. Greenness is calculated from satellite sensors that measure light reflectance from vegetation. Pixel values are the median day of year (2001-2017) the greenness of a pixel last reached 50% of the annual maximum (i.e. during senescence). Values greater than 365 indicate that the greendown occurred after Dec 31 and the date of maximum greenness occurred prior to Dec 31.

5.3 EVI Area Median

Source data: MODIS MCD12Q2 v006
Spatial resolution: 2.5 km
Temporal resolution: 17 year median
Temporal extent: 2001-2017
Units: Area under curve
Additional Features: none
Available Via: Images (LSP landing page); Geoserver Request Builder

Description: The total greenness that occurs between green-up and green-down is considered an important metric of productivity. Greenness is calculated from satellite sensors that measure light reflectance from vegetation. Pixel values are the median cumulative greenness (as measured by the 2-band Enhanced Vegetation Index) for an individual pixel. That is, summed daily values of EVI2 between greenup and greendown.

5.4 Mid Greenup MAD

Source data: MODIS MCD12Q2 v006
Spatial resolution: 2.5 km
Temporal resolution: 17 year MAD
Temporal extent: 2001-2017
Units: Days
Additional Features: none
Available Via: Images (LSP landing page); Geoserver Request Builder

Description: Median Absolute Deviation (MAD) of MidGreenup days of year for the period 2001-2017. The MAD is a robust alternative to the standard deviation that captures variability in data. It is calculated as the median of the annual anomalies from the median of MidGreenup over the period 2001-2017.

5.5 Mid Greendown MAD

Source data: MODIS MCD12Q2 v006
Spatial resolution: 2.5 km
Temporal resolution: 17 year MAD
Temporal extent: 2001-2017
Units: Days
Additional Features: none
Available Via: Images (LSP landing page); Geoserver Request Builder

Description: Median Absolute Deviation (MAD) of MidGreendown days of year for the period 2001-2017. It is calculated as the median of the annual anomalies from the median of MidGreendown over the period 2001-2017.

5.6 EVI Area MAD

Source data: MODIS MCD12Q2 v006
Spatial resolution: 2.5 km
Temporal resolution: 17 year MAD
Temporal extent: 2001-2017
Units: Area
Additional Features: none
Available Via: Images (LSP landing page); Geoserver Request Builder

Description: Median Absolute Deviation (MAD) of EVIarea for the period 2001-2017. It is calculated as the median of the annual anomalies from the median of EVIarea over the period 2001-2017.

5.7 Mid Greenup TS Slope

Source data: MODIS MCD12Q2 v006
Spatial resolution: 2.5 km
Temporal resolution: 17 year trend
Temporal extent: 2001-2017
Units: Days per Year
Additional Features: none
Available Via: Images (LSP landing page); Geoserver Request Builder

Description: Theil-Sen slope magnitude for pixels where the associated p-value is less than or equal to 0.05. It represents the change (days per year) of the MidGreenup value over the period 2001-2017. The Theil-Sen estimator is more resistant to outliers than linear regression equivalents. It is calculated as the median of the slopes of lines between all possible point pairs (i.e. day of year of MidGreenup and its calendar year).

5.8 Mid Greendown TS Slope

Source data: MODIS MCD12Q2 v006
Spatial resolution: 2.5 km
Temporal resolution: 17 year trend
Temporal extent: 2001-2017
Units: Days per Year
Additional Features: none
Available Via: Images (LSP landing page); Geoserver Request Builder

Description: Theil-Sen slope magnitude for pixels where the associated p-value is less than or equal to 0.05. It represents the change (days per year) of the MidGreendown value over the period 2001-2017. It is calculated as the median of the slopes of lines between all possible point pairs (i.e. day of year of MidGreendown and its calendar year).

5.9 EVI Area TS Slope

Source data: MODIS MCD12Q2 v006
Spatial resolution: 2.5 km
Temporal resolution: 17 year trend
Temporal extent: 2001-2017
Units: integrated EVI2 units per year
Additional Features: none
Available Via: Images (LSP landing page); Geoserver Request Builder

Description: Theil-Sen slope magnitude for pixels where the associated p-value is less than or equal to 0.05. It represents the change (integrated EVI2 units per year) of the EVIarea value over the period 2001-2017. It is calculated as the median of the slopes of lines between all possible point pairs (i.e. EVIarea and its calendar year).