Gro Yield Model Forum
Welcome to Gro’s Yield Model Forum! Please feel free to post any comments, feedback, or questions here. Note that you can opt to receive email notifications to stay updated on the threads you comment by going into your user settings.
Check out our yield model page to learn more about the model and download our paper: https://gro-intelligence.com/yield-model
54 results found
-
1 vote
State and county level results will be made available to users of the Gro platform, but are not on our public website.
-
1 vote
We intend to use any data that’s available and useful. We made heavy use of ground-based data in the US model because it’s the prototype for all of the other models that we are building in areas with poor ground-based data. We believe that satellite data alone will allow us to successfully model most crops around the world now that we have the US results.
-
What is the effect of higher volatility due to climate change?
Your models will get better with time as it "learns", but climate change makes it that the past is not a good predictor of the future.
1 voteClimate change occurs slowly enough that the refitting of the model with each year’s new data captures the effect pretty well. In a case of discontinuous rapid climate change, our model would suffer from poor results due to its historical bias.
-
1 vote
Clearly, forecasts with high confidence have to wait for greenness and other data that only arrives after the crop is growing. Nevertheless, our model does have statistical value as early as the beginning of May.
-
1 vote
-
Weekly Commentary - June 13, 2017
We’re a little relieved to see the model yield rising from 154.79 bushels/acre (or 9.72 tonnes/hectare) to a still-alarming 156.74 (or 9.84), an increase of almost 2 bushels/acre. Of course, this is still well below trend and consensus forecast yields, and if realized would be hugely bullish for corn. The late-May increase in land surface temperatures explains most of the move in the model, which occurred despite percent good-to-excellent deteriorating by 1 point in Monday’s Crop Progress report. Since we also see in that report that 95 percent of the US corn crop has emerged from the soil, we expect…
1 vote -
Weekly Commentary - June 5, 2017
Our model value has crashed to 154.79 bushels/acre (or 9.72 tonnes/hectare) from 162.57 bushels/acre (or 10.20 tonnes/hectare). There are two big factors at work here: the lower-than expected corn crop conditions estimate in USDA's Crop Progress reports so far and colder than normal temperatures in late May causing a big drop in forecast Indiana/Eastern Corn Belt yields.
As data has accumulated, our confidence in a lower-than-expected yield has grown. Corn plants are emerging in the Midwest and satellite-based normalized difference vegetation index (NDVI) information will soon begin to influence the model and increase its accuracy. At this early date, we…
1 vote -
Weekly Commentary - May 28, 2017
This week, we see another wild swing in the yield model, clearly caused by an extreme change in average temperatures across the US Midwest. The value has fallen from 168.60 bushels/acre (or 10.59 tonnes/hectare) to 162.57 bushels/acre (or 10.20 tonnes/hectare), which is only slightly above the alarming initial value in early May. It’s important to remember that early in the season, the model has less data and therefore can fluctuate dramatically with changing conditions. But with recorded temperatures this far below average across a wide geographic area, the US corn crop needs seasonal warming to occur soon to reach its…
1 vote -
2 votes
The current yield model is for the US – it was modeled at the county level and extrapolated to the national level with 15 years of backtested data. However, we are planning to calculate yield projections for a wider variety of crops and regions. In the meantime, you can access the data used in the model at a global scale in Gro, our global agricultural data platform.
-
Weekly Commentary - May 20, 2017
The yield model value has risen from 167.85 to 168.60 bushels/acre, and appears to be converging to trend, which is currently at 169.46. The forecast has gotten much less bullish for corn in the past two weeks, largely due to a land-surface temperature catch-up that’s still taking place. As expected, the week-to-week variation looks more stable since more data came in. Note that the model yield is still bullish, since it’s below the trend and below the trade estimate/guess at 170 bushels/acre.
1 vote -
Weather Forecasts?
Does your model use forward-looking weather forecast data, or only observed/actual values?
2 votesOur model currently only uses remotely sensed weather data, specifically LST. We are also testing other weather data now. Let us know if you have any feedback or suggestions!
-
4 votes
We don’t at the moment. The model is modeled at the county level so we do plan to make county level forecasts available to the public in the near future.
-
Is there a way to go back and see the output from previous years on specific dates during that year?
For example if I wanted to see what Gro was forecasting as the yield on July 10, 2016 or August 1 2015 etc.?
3 votesYou can see the output from previous years in the paper (i.e. figure 4, figure 7, figure 9, etc.). You can download the paper here: https://gro-intelligence.com/yield-model
-
3 votes
- Don't see your idea?