> For the complete documentation index, see [llms.txt](https://eflow.gitbook.io/ffc-readme/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://eflow.gitbook.io/ffc-readme/functional-flow-calculator/metrics/spring-recession/rate-of-change.md).

# Rate of Change

## Definition

The spring recession rate of change is the median daily rate of change in flow from the start date of the spring recession until the start of the dry season low flow season. Only days with negative change (i.e. a decreasing rate) are used in the calculation. This metric is measured as a percentage rate of decrease.

## Steps

1. Calculate the spring recession timing and dry season low flow timing. Rate of change is calculated between these two timing indices.
2. For each water year, check if both the spring recession and dry season timing dates exist, and make sure the spring date occurs before the dry season date:

   ```python
   if not math.isnan(spring_timing) and not math.isnan(summer_timing) and summer_timing > spring_timing:
   ```
3. For each daily time interval, calculate a simple daily rate of change, where change is equal to (Day2 - Day1)/Day1. Rate of change is only recorded for those dates in which the daily change is negative.

   ```python
   flow_data = raw_flow[int(spring_timing) : int(summer_timing)]
   for flow_index, data in enumerate(flow_data):
       elif flow_data[flow_index + 1] < flow_data[flow_index]:
          rate_of_change_neg.append((flow_data[flow_index] - flow_data[flow_index + 1]) / flow_data[flow_index])
   rocs_only_neg.append(np.nanmedian(rate_of_change_neg))
   ```