The Oregon State University Oregon IPM Center, in cooperation with The National Plant Diagnostic Network
(NPDN), USDA NIFA (Western and other Regional IPM Centers, NRI/AFRI Plant Biosecurity, EIP, CPPM ARDP, Cooperative Extension Service, others), the IPM PIPE programs and USDA RMA, APHIS PPQ, numerous grower-supported weather networks,
Oregon Climate Service, the OSU PRISM Group, the Western Weather Systems Workgroup, various regional Climate Hubs, the USGS National Phenology Network, other Universities, NOAA, NWS, NRCS, the US Bureau of Reclamation, and others,
has been working to make real-time daily weather data, degree-days, forecasts, pest and plant disease risk models and maps available to serve
plant biosecurity, agricultural and pest management needs for the US and SW Canada. These formatted weather data, degree-days, and
pest model products are kept current and relevant for
decision support for Extension personnel, growers and farm managers, consultants, invasive species surveys, regulatory needs, researchers, and students. Please
take 2-5 minutes and fill out our User Survey and give us feedback so we may better meet your particular needs.
LOG OF WHATS NEW - 2010 to 2015
For extended forecasting, (formerly based on 30-year Normals), began using 10-year average station data, adjusted using PRISM-CAI (climatologically aided interpolation) as the default for our standard degree-day calculation products. To determine if this is true for the particular toolset you use, with full (daily) output, you should see, after any short term (5-7 day) forecast, a line that reads "Forecast using: 10 yr Hist (2004-2013) Avg StaCode PRISM-CAI est", where StaCode is the station currently selected (assuring that it is relevant to this site). We expect to update these 10-year average data every 2 years so they are kept relatively current. Also note that we are working to make a suite of options available for extended forecasts in the next 1+ year at this website.
Added 3 new degree-day models: 1) Pink Bollworm, a long-time pest of cotton. By request of APHIS-PPQ to support on-going trapping programs. 2) & 3) Arcadia and Green Magic Broccoli vegetable models, for the CROPTIME project.
Changed old degree-day modeling "ddmodel.pl" interface to redirect to newer "ddmodel.us" interface, as the older interface is no longer being supported and may fail to work properly. The newer interface mainly differs in using Google maps to select the weather station. We had not been using the old interface for 2 years so it is time to let it retire...please change your bookmarks accordingly. FYI if you require the old "ddmodel.pl" for some reason, it was renamed to ddmodel2.pl, as it is still being used as the "middleware" program that passes parameters onto the DD calculator.
Added Harcourt 1981 model for alfalfa weevil, by request of U. WY. We expect to add a more robust model that can be started Jan 1. The Harcourt model was developed in Ontario and has a start date of Mar. 1st, so is unlikely to be useful in warmer states e.g. Arkansas.
Revised the "simple average DD" calculation formula (by request) to use a horizontal cutoff for when the daily maximum temperature is above an upper threshold if provided. This will limit degree-days both for the site and mapping DD calculators.
Updated station and network database to include over 4,000 new stations, making the approximate total 25,400 weather stations, up from ca. 20,500 for 2014. These new stations were inadvertantly missed when the station database was updated last December, and this will result in outages for Jan and Feb for the new stations. We hope this will not pose a major problem as the growing season is just getting started. New models for experimental use and testing include mummy berry of blueberry and dollar spot of turfgrass. Added new slide show (pdf) Crops and Climate - Has it been getting warmer in the Pacific NW & how will that affect plant/crop phenology? from the Small Farms Conference, Corvallis OR Feb 28, 2015.
Added model for Japanese beetle based on USDA APHIS trapping data, requested by Oregon Dept. Agriculture. The model would benefit from better lab data on temperature development, and by including more trapping data, so consider this model "provisional" until further updates are available.
Updated model for emerald ash borer based on recent M.S. thesis by SAD Duarte, Ohio State Univ. 2013. The study looked at adult emergence in six locations spanning Ohio to S. Michigan for the years 2011-2012. The previous version of the model used here had a starting date of Mar. 1. The new model and a model used by APHIS PPQ use a Jan 1 start date, which may be more inclusive of southern states.
Updated Degree-Day Mapmaking to include "Network Selection" feature that allows users to control the types of weather networks used in the process of making degree-day maps. In the past we have always started with potentially all weather stations to influence DD map results. We are starting with 3 options, first by including fewest and best networks only as the default. Second option is to exclude certain networks that tend to have poorly cited and maintained weather stations. Third option is to include all networks; as the tool was configured up until now. This setting can influence several outcomes in the DD map, including 1) Frequency of possible anomalies including "bulls eyes" caused by stations not calibrated or sited the same as their neighbors, 2) Local map accuracy (exclude a network with a good performing weather station in your area, and your local results may be less accurate, and 3) Speed of map production (fewer stations require less process time). We also have continued to improve the QA filtering process used in map production, and attempt to determine which stations should not be used based on the amount of missing and out-of-range data. Also another server (fusar) was configured to make DD maps. Try this server if the main one is slow or not working at the moment.
Added a new "model preview" feature to the main interface to our DD models, ddmodel.us. This allows seeing model output (default now set at 3 events; user selectable) and an additional weather station QA score everytime you change any settings. This new feature will greatly speed up the use and add to ease of use of phenology models. Several other bug fixes and features have been integrated into this system, hopefully making it much more accurate and user-friendly for most people.
A number of updates have been implemented for the 2014 season, including: 1) Improvements to the main interface to our DD models, ddmodel.us. These include bug fixes and a model category for pome fruits. 2) Updates to our daily max/min temperature quality assurance system. Now in addition to using PRISM CAI (Px/Pn tags), we use our max/min data from our hourly virtual weather data (Vx/Vn tags), and site-tied 30-year historical weather (Hx/Hn tags), and as a last resort, the older time averaging method (Mx/Mn tags). These new methods will help in finding and using better weather stations and data. 3) Repairs to our disease risk grids for late blight of potatoes. 4) A new version of the grass seed stem rust decision support system. 5) Degree-day model outputs include the NWS NDFD 7-day forecast for most applications as a default. Older NWS zone forecasts are still used for some applications.
Updated our disease mapping system to allow 5-hour, 10-hour, and 3 day, every 3-hour loops in addition to the 5-day loop. This allows for much greater precision in visualizing disease risk developments and in the weather parameters that drive the disease risk models. Just configure your region, weather parameters, and plant disease risk models, and click on the "Submit" button to allow displaying any of these elements on the google map on left. The "Loop" buttons are just above the Google map.
Developed with the help of OSU Media Communications.
Fixed a bug in ddmodel.us so that now if you change models the start date will be reset to the correct date IF the model uses a calendar-based starting date (such as Jan 1 or Mar 1). However, if you change models to one with a non-calendar based starting date, then the starting date does NOT reset, since it is assumed you need to enter your starting date based on sampling information (e.g. BIOFIX date for codling moth, based on first consistent capture in pheromone traps). Please, always check all inputs before running any model, otherwise outputs may be invalid!
Updated the new tool ddmodel.us that provides a better user interface making use of Google maps for location selection, and fewer options to simplify use. There are still glitches that are hopefully minor. Overall, this should be a major improvement over "ddmodele.pl" which this new tool is actually a front-end of. New features added since last version: 1) IP location finder (like the "MyPest Page" interface, finds nearest weather station to your internet provider, showing your location in the Google Map), 2) new page proportions should work better on small devices, 3) hidden threshold and calculation method whenever a model is selected, 4) more extensive report on model validation status and regions where model is known to be in use, 5) other minor changes. Let us know if you would like other features added to the system!
New models added recently for boxwood blight (invasive disease of boxwood), Galerucella calmariensis, a beetle used for biocontrol of purple loosestrife, and Asian citrus psyllid, a pest and vector of citrus greening disease of citrus. 101 models total to date (plus the separate grass seed stem rust and swiss needle cast models). Forecast grids recently added to disease risk mapping regions and models, as supported by a recent USDA NIFA AFRI project.
Posted a new (beta - not finished but working fairly well) Google Maps USA interface to degree-day models and calculations. This new interface has several improvements and can grow to accomodate others: 1) Uses Google Maps for location selection (so now serves as an interface for all stations in our database (US Canada Mexico Virgin Islands etc), not just Pacific NW States), 2) New model categorization making model selection (currently 78 available) easier, 3) Automated selection of nearest forecast zone and historical weather data based on location selected, and 4) ability to display custom inputs for different categories of models. For example, selecting "plant diseases" will display a link to "MyPest Page" for access to additional hourly weather driven plant disease risk models. This interface requires some more work to be smooth, for example your inputs such as starting year currently will reset when you select a different location. Also, the options for "Table", "Graph", "DDStat report", etc. do not work as they should. Currently the default, full output options are used with no ability to suppress them. Feedback is always welcome!
Posted a table of ddmodel usage from the log files of number of degree-day model and calculator runs by model and by year. In short, model usage has grown from 5,714 runs in 2000 and 13,020 runs in 2005 to 47,795 runs in 2011 and thus far in 2012, 75,844 runs.
Updated degree-day mapmaking to use new 30-year normal PRISM data (1981-2010), thanks to the OSU PRISM Group. This will make a rather minor difference in DD maps, but it should be an improvement.
Updated the "ddmodel.pl" degree-day/phenology modeling program. It will give new reports using the "DD Clock" feature (see 6/15/2012 below), and use the same jquery "fancybox" popup for summary table info (click on graph) rather than the old type of popup (that appeared on hover). Hopefully these new features will work on most all major browsers. The new DD Clock info takes about 2-3 seconds longer to calculate, but we think the extra info should be worth it.
Added a new model for light brown apple moth, Epiphyas postvittana, an invasive species that feeds on many plants, originally from Australia. The model is very likely to change as new data become available.
Added a new spotted wing Drosophila IPM PIPE, to be a part of the Western Specialty Crops PIPE. This site has several innovations, including an overwinter mortality model, historical phenology maps, and custom daily phenology location tables and Google maps with "DD Clock", showing for a given location the DDs from the map, DDs for the nearest weather station, and an estimate of how many days the station is ahead or behind for the current year compared to the last two years and to the 30-year normals. This feature was also added to the standard DD maps such as the Google Map for base 41 US DDs. The new SWD page is linked from the "Shortcut Links" tab on the home page (middle of page). Other decision support tools for SWD are planned for the site in the near future.
Updated DD mapmaker for 2012. The program now accesses 1981-2010 30-year Normals.
Added a model for omnivorous leaftier, Cnephasia longana, a pest of nursery crops, small fruit, hazelnut, hop, and other crops, at the request of a private consultant working in the Willamette Valley, Oregon. This model is based on recent sampling of larval infestations, plus historical (1930-1951) OSU studies of biology and life history. As with any new model, use with caution and let us know if you think it needs improvement.
Recently added models including a new wheat model (Bauer model for Spring wheat) requested by Wyoming Extension Service, an experimental model for overwintering mortality of spotted wing Drosophila (tricky to use as it starts Nov. 15 of previous year), based on laboratory research and field trapping data, and Fender's Blue Butterfly, a threatened/endangered native species of the Willamette Valley, Oregon (in cooperation with numerous agencies including the Nature Conservancy, The Xerces Society, The U.S. Fish and Wildlife Service, U.S. BLM, U.S. Corp of Engineers, and Benton County, Oregon). This latter model should help conserve sampling efforts by helping to predict flight times, and aid in protection of this species and its habitat.
Most resources have been updated for the new year (except the custom DD mapmaker program is pending an update). Historical normal max-min temperature data have been updated to 1981-2010 normals from NCDC public data. Most online resources now reflect this update (DD mapmaker is an exception).
At the request of the Director of WSU AgWeatherNet network, we have removed display of max/min/precip data normally displayed with degree-day model outputs, and removed weather data from chart pop-up table summaries in the risk/models program. We are sorry about any inconvenience that this may cause. Normally our policy is that the data should be visible to assist users in assessing weather data quality. In the case of AgWeatherNet, there should not normally be a problem with lack of data display because there have been few outages or data QA problems with our feed of the data. However, we urge extra caution in using any model in which you cannot see the weather data inputs. You may wish to join the WSU AgWeatherNet network for access to data directly from their website.
Added a model for Douglas-fir needle midge, Contarinia spp. phenology, by request of christmas tree producers in Oregon. This model is provisional, and is a synthesis of published data from 1990 and recent (2009-11) trap data, all from W. Oregon. We determined the starting date (Jan. 1) and lower threshold (Tlow=39F) for development using the "lowest C.V." method, as laboratory rearing data have not been identified at this time. Use the model with caution, as it is likely to change as more data are discovered. Note that this model differs significantly from other Extension recommendations, which use Tlow=50F, with 200-400 DD and a Mar. 1 start date (Mich. State) and 270 DD and Jan. 1 (Penn State) for beginning adult emergence. Our lowest C.V. analysis indicates that a Tlow of 50 is too high to be used for this insect.
Added a model for pine shoot beetle, Tomicus piniperda L. swarming activity, by request of USDA-APHIS. This model is a re-working of a model published from Norway and Finland data by Saarenmaa (1989), who showed that swarming depends on 2 or more days of maximum temperatures greater than a threshold, which is set at 54F for our treatment of the model. Also, to support the model, we have added "Tmax Heat Units", where temperature is accumulated on days when the Tmax is greater than a lower threshold. If an upper threshold is used (set to 70F for pine shoot beetle), then the Tmax is reset to the upper threshold on days when Tmax exceeds the upper threshold (thus limiting the response to a maximum on hot or warm days). This type of heat unit was added to the degree-day calculation tool and to the degree-day mapmaking tool. As many insects begin mating and reproduction based on a threshold that is compared to the daily maximum temperature, rather than a degree-day accumulation, this new calculation method should be useful for other organisms. Use the pine shoot beetle model with caution, as it has not been tested in the US yet.
Added a model for western flower thrips, by request. This model is a new synthesis from numerous sources and may give an indication of generation time under differing climates. Use with caution. Subject to change.
Updated usmapmaker.pl to the new server, with several major bug fixes and a new interpolation algorithm. On the new server, mapmaking should be ca. twice as fast as on the old server. See the "Degree-Day Maps" tab for links to using the service on any of 3 servers.
Details of changes: updated all links to proper year(s), revised legend creation, changed inverse-distance-weighted interpolation (s.surf.idw) to regularized-spline-with-tension (s.surf.rst), because the former is not working in 64-bit. Note that the daily DD maps are created on the older server, where the idw interpolation method remains the method used. The rst method is superior overall, but can produce a greater number of undesirable anomolies.
Revised W. cherry fruit fly DD model to better serve Mid-Columbia Region of Oregon. Also, the DD calculation method was changed from Simple Average to Single Sine, which for 6 sites in the Mid-Columbia Region, results in predictions that are on average ca. 3 days earlier than before.
Switched to a new web applications server, an HP DL 380 G7 w/24GB memory, 64bit Linux, etc. Due to the change from 32 to 64 bits, we are redirecting all GrassLinks applications to one of our 2 other servers. Several issues remain to be fixed, but most services are working. Overall the new server should be around 3-times faster, and handle much higher loads than the old one.
Added brown marmorated stink bug DD model to database. IPPC synthesis from Nielsen et al. (2008, 2009) data. This model is preliminary as it is based on limited lab + field data primarily from 1 location (Allenstown PA). It may perform well in areas with a similar climate, but beware for cooler and especially for warmer climates.
Also, we began to categorize hourly driven models, now that there are 18 of them. The new models include Melcast for muskmelon and Melcast for watermelon, and strawberry powdery mildew. These are currently being verified.
Updated European grapevine moth model in collaboration with UC Cooperative Extension. The DD (F) requirements for first flight were changed slightly (from 190 to 270 DD), for first egg laying (from 390 to 450 DD), for 1st egg hatch (from 510 to 540 DD), and for 95% flight (from 680 to 720 DD) to better match results seen in Napa and Sonoma counties, 2010-2011.
A new feature was added to the risk models page that attempts to fill in missing and out of range temperature, dewpoint and other weather variables using virtual weather, that is estimated from nearby stations. This feature can be switched off (and other QA flagged data substitutions turned on in the settings portion of the page).
Added new Google Maps interface to year-to-date degree-day maps. While not as feature rich as the GRASSLinks interface to the same maps, you are able to query within the maps for specific location degree-day values, which is not a standard feature provided by Google Maps. These links are provided under the thumbnails for each region, e.g. see US DD Maps Index Page. Also we will begin to refer to the entire website as "MyPest Page" as a shortcut for the byline, "IPM Pest and Plant Disease Models and Forecasting for Agricultural, Pest Management, and Plant Biosecurity Decision Support in the US".
We are beginning to use virtual weather data to fill in missing and other problematic hourly weather data, similar to what we have been doing for daily Max/Min temperature values for years. This should help with QA/QC of weather data and pest models, and improve our ability to use more models under more circumstances. Soon this will be an option that end-users will be responsible for setting up for their own needs in the display settings area. More on this soon...
Added new model for Early blight of potato to degree-day models database. This model is based on Sands P-Days type of degree-day calculation. P-Days can also be generally used for potato phenology. Also added Sunflower Stem Weevil model.
LOG OF WHATS NEW - 2010
Added new model for European grapevine moth/Lobesia botrana to degree-day models database. This model is based on available world literature and UC Napa Cooperative Extension data. The model is presumptive and experimental. It differs from the mini pest risk assessment (APHIS) model primarily in the biofix information, which was matched to Napa CA 2010 only. Thus the model may soon be revised as more data become available.
Popup tables added to weather parameter charts in /risk/models (similar to those for diseases and DD models). Click on weather station pin in google maps to reset display to the new station (allows better smart phone support).
10/6/2010 New features added to /risk/models "MyPest Page", including popup tables for disease charts (similar to those for DD models), a larger Refresh button to help remind users to use it, now disease charts display the year in the chart, bug fixes for several models, and updates to degree-day calculator.
Re-organized home page to a tabbed version. Try the "Quick Start" tab - enter your crops and zipcode, to
run the risk models program with defaults at your location. risk/models has been revised to use these crop-based defaults; plus the weather data is now displayed above any pest and disease models. These changes should help new users of the system get started.
New features added to the risk/models "MyPest Page": 1. Integrated degree-day charts on page. Click on these charts to display a table version of the same DD summary results. Above the table, there is weather data QA report and link to "Show daily output in new window..." which allows you to run the full output in a new window. 2. Now defaults are being defined for 1st time users of the page.
Revised model for spotted wing Drosophila, described here. This model is very likely to evolve rapidly, as this is a new pest whose springtime activities are now being monitored. Updated degree-day/phenology models with ability to run models from this listing/summary table.
Updated the main risk models page, uspest.org/risk/models as our first release of "MyPest Page" that adds several new features to risk models:
1) Integration of the full database of degree-day models (which use daily max/min temperatures rather than hourly data so results are displayed in a separate browser window).
Note that currently all model settings must be saved by clicking the "Refresh" button before running any DD models.
Using "Refresh" will also set cookies on your computer, so that next time you visit the page all settings will be preserved.
Also this allows you to run your selected DD models from nearby stations using Google Maps
(hover the mouse over a station pin to see the model links); and
2) Two types of chilling unit models: 1. Simple/user configurable and 2. Utah (more complex but with fixed settings); these are used to predict vernalization requirements for many perennial fruit and nut crops.
3) Option of NWS Digital Forecasts (second option under Settings);
Updated the degree-day mapmaker to use higher resolution PRISM data (800m), to better support subregions for faster processing, and to allow users to download the data for thier own GIS processing and mapmaking needs. Also updated documentation for usmapmaker.pl.
In addition, because the FLLW algorithm does not account for rainfall-induced leaf wetness, the Fox Weather LLC/IPPC weather/plant disease modeling system includes a precip-drying algorithm (FoxLW). This presumes an exponential-based drying curve and takes into account windspeed, seasonal adjustment, and day/night time.
Details of the FoxLW addition are:
-For rainfall at 0.05 in/hr, LW=10 for that hr
-at 0.02 in/hr: in summer LW=5, winter or
at night then LW=10.
-LW should end < 0.5 hr after rain ends
during day, and about 1 hr after rain
ends at night if WS=5mph.
-if WS=10mph, then LW ends when rain ends.
-NO LW for rain < 0.02 inch with WS >= 10mph.
-LW = 3 if rain > 0.01 or < 0.02 with WS <=5mph
-WS ranges from 5 to 10mph, LW 3 to 0.
The Kim et al. FLLW and FoxLW models are independent and parallel computations for a Leaf Wetness value: We output the HIGHEST LW of the two (they are NOT ADDED together, we just take the MAXIMUM value).
Changes and updates to this system are in development and expected in forthcoming releases.
No claims are made as to the correctness or appropriateness of this information for your particular needs. No specific pest control products are intended for
endorsement or use. All responsibility rests solely with the people who interpret and implement information from this and other sources. Use this information
with caution and at your own risk - errors occur, and predictive models do not replace the need for proper monitoring in the field. If you observe conditions that differ substantially from
model predictions, please contact us to determine if the model inputs were incorrect, if the model functioning or weather data are in error, or if the model is inappropriate
for your conditions.