10.4. Estimation Techniques

Arithmetic Mean Method

Applicability: When rainfall at surrounding stations is fairly uniform (variation <10%).

Formula:

$P = \frac{1}{n} \sum_{i=1}^{n} P_i$

Where:

• $P$ = missing rainfall value at station X
• $P_i$ = rainfall at ith neighboring station
• $n$ = number of neighboring stations

Advantages:

• Simple and quick.
• Good for flat terrains.

Limitations:

• Inaccurate in non-uniform orographic regions.

The Arithmetic Mean Method is used when the rainfall values from surrounding stations are relatively close, specifically when the variation in rainfall is less than 10%. This method averages the rainfall data from neighboring stations to estimate the missing data at a specific station. The formula involves summing the rainfall amounts from all nearby stations and dividing by the number of stations to find the average. While this method is easy to apply and works well in flat areas, it becomes unreliable in regions with significant differences in rainfall due to geographic features.

Imagine you have five friends who live nearby, and you want to estimate the total amount of rain they each received on a day when you forgot to record yours. If they all report similar amounts, it's reasonable to say you likely had a similar experience. However, if one of your friends lives on a hilltop and reported a much higher amount, relying on the average would misrepresent your own situation.

Normal Ratio Method

Applicability: When normal rainfall at surrounding stations differs by more than 10% from the station with missing data.

Formula:

$P = \left( \frac{N_x}{N} \right) \sum_{i=1}^{n} P_i$

Where:

• $P$ = estimated rainfall at station X
• $N_x$ = normal rainfall at station X
• $N$ = normal rainfall at station i
• $P_i$ = observed rainfall at station i

Advantages:

• Adjusts for climatic variability.

Limitations:

• Requires long-term normals.

The Normal Ratio Method is applied when the normal rainfall amounts at stations around the missing data station vary significantly, that is more than 10%. This method adjusts the observed rainfall data based on the historical average rainfall (known as normals) at the missing station relative to its neighbors. It involves calculating a ratio of the normal rainfall at the target station to the normals of surrounding stations, and then applying this ratio to the observed rainfalls to generate an estimate. Though effective, this method requires good historical data over a long period to determine these normals accurately.

Think of it like adjusting a recipe. If you're baking cookies and your friend's oven is set to a different temperature (as their normal), it would make sense to modify the cooking time according to that difference when sharing the recipe or serving the cookies to ensure they're cooked just right.

Inverse Distance Weighting Method (IDW)

Applicability: When distances between the station with missing data and nearby stations are available.

Formula:

$P = \frac{\sum_{i=1}^{n} \frac{P_i}{d^2_i}}{\sum_{i=1}^{n} \frac{1}{d^2_i}}$

Where:

• $P$ = estimated rainfall at station X
• $P_i$ = rainfall at station i
• $d_i$ = distance between station i and station X

Advantages:

• Considers geographical proximity.

Limitations:

• Requires accurate distance data.
• May not work well in regions with varying topography.

The Inverse Distance Weighting Method (IDW) estimates missing rainfall data by considering how distance affects rainfall amounts; data from stations closer to the missing data location are weighted more heavily than those further away. The method involves dividing the observed rainfall data from each nearby station by the square of its distance from the station with missing data, summing these values, and normalizing them against the sum of the weights. This increases the accuracy of the estimate by acknowledging that closer stations provide more relevant data. However, the method relies on having accurate distance measurements and might be less effective in areas with varied terrain where rainfall patterns can change dramatically over short distances.

Imagine you lose your phone at a party, but you notice your friends who are closer to the location where you lost it start looking for it right away, while those on the other side of the room take longer. It makes sense that the ones closer would have a better idea of where to find it. Similarly, the IDW method gives more weight to nearby rainfall stations when making estimations.

Multiple Regression Method

Applicability: When relationships between rainfall at different stations are linear and statistically significant.

General Form:

$P = a_1 P_1 + a_2 P_2 + ... + a_n P_n + C$

Where:

• $P_1, P_2, \ldots, P_n$ are rainfall values at neighboring stations
• $a_1, a_2, \ldots, a_n, C$ are regression coefficients determined statistically.

Procedure:

1. Collect data from nearby stations.
2. Conduct regression analysis.
3. Apply regression equation for estimation.

Advantages:

• Highly accurate if data is consistent.
• Captures correlation among stations.

Limitations:

• Requires computational tools.
• Sensitive to outliers.

The Multiple Regression Method is a statistical approach used when there is a clear linear relationship between the rainfall amounts at different stations. In this method, data from multiple neighboring stations is collected, and a regression analysis is performed to determine coefficients that represent the relationship between the rainfall at the target station and those at neighboring locations. The resulting regression equation can then predict the missing data point. This method is highly accurate when the data is reliable and can account for interactions between multiple stations but necessitates the use of statistical software and can be affected by extreme data points (outliers).

Consider a scenario where a student gets grades in different subjects, and you want to predict their possible score in math based on their scores in science and English. By examining how scores in those subjects relate to each other, you can use a regression model to make an educated guess. Similarly, in rainfall analysis, this method understands how data from various weather stations correlates to better estimate missing values.

--